10484.pdf
Media
- extracted text
-
WORKING PAPER -
Industrial Tariffs and
South Asia
Interpreting for Development
Clelnltlald
Centre for Trade & Development
t’; nF? i
Industrial Tariffs and
South Asia
Interpreting for Development
Prabhash Ranjan
Research Officer
Centre for Trade and Development (Centad)
F
Copyright © Centad May 2006
Ccntad Working Papers are intended to disseminate the preliminary findings of
ongoing research both within and outside Centad on issues around trade and
development for the purpose of exchanging ideas and catalyzing debate. The views,
analysis and conclusions are of the authors only and may not necessarily reflect the
views or position of Centad or Oxfam GB. Readers are encouraged to quote or cite
this paper with due acknowledgement to the author and Centad.
Centad and the author are grateful to S. Narayanan, B L Das, Mustafizur Rahman,
Darlan Fonseca, Biplove Choudhary and an anonymous referee for providing
useful comments on a shorter version of this paper and to Palakh Jain for providing
valuable research assistance. Views and errors if any are solely author’s responsibility.
This paper draws from author’s paper ‘Choosing the Appropriate Tariff Reduction
Formula in NAMA’, XLI (16), Economic and Political Weekly and ‘Agriculture and
NAMA Negotiations: Searching for the Landing Zone’, Working Paper 4, Centre
for Trade and Development (Centad) (2006).
Key words: NAMA, South Asia, Swiss formula, WTO
Design and Printing by
New Concept Information Systems Pvt. Ltd.
Plot No. 5, Institutional Area, Sarita Vihar
New Delhi- 110 076
1U ;t 3 4 I
Foreword
The WTO negotiations on industrial tariffs, commonly referred to as Non Agriculture Market Access
(NAMA) in the World Trade Organisation (WTO) seem to be acquiring a dangerous direction. Developed
countries have been applying unreasonable pressure on developing countries to undertake very steep
reductions in their industrial tariffs. On the face of it, this is a clear attempt to deny developing counties
of an important economic and developmental tool of industrial tariffs forever. It is important to recall that
the present developed world allowed its industry and manufacturing to blossom behind high tariff walls
for decades.
The developments in Geneva seem to indicate that the pressure of developed countries is perhaps working.
A consensus seems to be emerging for a ‘Swiss formula with two coefficients’ that would lead to drastic
cuts with higher tariffs being subject to deeper cuts. Steep and drastic cuts in industrial tariffs would erode
the existing policy space in developing countries and may subject many nascent industries to international
volatility. Apart from adverse economic consequences, this may also have undesirable social consequences.
The irony is that although the developed world expects developing countries to undertake drastic cuts, it
is not at all forthcoming in dealing with the issue of tariff peaks and tariff escalation. Under the influence
and pressure of huge domestic lobbies many developed countries maintain extremely high tariffs especially
on products of export interest to developing countries. This is clearly special and differential treatment
working in reverse!
Given this context this paper moots a bold approach on the core issue of tariff reduction in NAMA. It very ably
builds a case for a tariff reduction formula, different from the ‘Swiss formula with two coefficients’ using the
declaration adopted at the Hong Kong ministerial conference in December 2005. The formula proposed in this
paper argues that the only way principles like less than full reciprocity (LTFR) and elimination of tariff peaks and
tariff escalation could be honoured is by having the average tariff rate of a country as one of the coefficients in
the formula. The paper has used the industrial tariff profile of India and Pakistan to show the efficacy of such a
tariff reduction formula. It is important to recall that the principle of LTFR is an integral part of the Doha Work
Programme, July 2004 Framework Agreement and the Hong Kong Ministerial Declaration. The paper suggests
that negotiators of developing countries could very well use the declaration adopted at Hong Kong to argue for a
formula that takes into account the average tariff rate of a country or other factors such as dependence on tariffs
for revenue, in addition to meaningful flexibilities.
Centad urges that the negotiators of developing countries including India and Pakistan to argue for such
an interpretation in their own as well as in the interest of other developing countries. In the meanwhile,
Centad is committed to do further research on this issue.
Dr. Samar Verma
Senior Policy Advisor and Trade Team Leader
Oxfam GB
Oxford
Abbreviations
ABI
Argentina, Brazil and India
BTR
Bound Tariff Rate
DDA
Doha Development Agenda
EC
European Commission
EU
European Union
GATT
General Agreement on Tariffs and Trade
HK
Hong Kong
HS
Harmonised System
LDCs
Least Developed Countries
LTFR
Less Than Full Reciprocity
NAMA
Non Agriculture Market Access
NGMA
Negotiating Group on Market Access
T&C
Textiles and Clothing
US
United States
WTO
World Trade Organisation
Contents
Foreword
Abbreviations
Executive Summary
Hi
io
vii
1.
Introduction
1
2.
NAMA Negotiations
3
3.
Scope of the Paper
4
4.
General Comment on NAMA Outcome of HK
4
5.
Swiss Formula with Coefficients
5
Possible Interpretations of Paragraph 14
Swiss Formula with Two Coefficients
ABI Formula
Carribean Formula
Less Than Full Reciprocity (LTFR)
Implementing Paragraph 14
Two Approaches
6
7
8
9
11
12
14
6.
Eliminating Tariff Peaks and Tariff Escalation
17
7.
Implementing the Tariff Reduction
17
8.
Conclusion
19
5.1
5.2
5.3
5.4
5.5
5.6
5.7
Annexures
20
Annexure 1: Impact of the ‘Swiss Formula with Two Coefficients’ on the Bound Tariff Rates
of Seven Sectors of India
20
Annexure 2: Impact of the ABI Formula on the Bound Tariff Rates of Seven Sectors of India
22
Annexures 3: Impact of the Caribbean Formula on the Bound Tariff Rates of Seven Sectors
of India
24
List of Tables
1
2
2B
3A
A
Industrial Tariff Profile of South Asian Countries
3
Impact of the Pure Swiss Formula on the Bound Tariff Rates of 5107 Tariff Lines of India
(HS 6 Digit Level) (Percentage)
7
Impact of the Pure Swiss Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistan
(HS 8 Digit Level) (Percentage)
8
Impact of the ABI Formula on the Bound Tariff Rates of 5107 Tariff Lines of Indian Industrial
Products at HS 6 Digit Level (Percentage), the Average Tariff Rate = 34.3 Percent (Percentage) 9
3B
4A
4B
5
6A
6B
7A
7B
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
Impact of the ABI Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistani
Industrial Products at HS 8 Digit Level (Percentage), the Average Tariff Rate = 35.3 Percent
(Percentage)
10
Impact of the Caribbean Formula on the Bound Tariff Rates of 5107 Tariff Lines of Indian
Industrial Products at HS 6 Digit Level (Percentage), the Average Tariff Rate = 34.3 Percent,
B = 1 (Percentage)
10
Impact of the Caribbean Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistani
Industrial Products at HS 8 Digit Level (Percentage), the Average Tariff Rate = 35.3 Percent,
B = 1 (Percentage)
11
Demonstration of LTFR (Percentage)
12
Demonstration of the Levels at which the Coefficients should be to Honour Real ‘LTFR',
Developed Country (Percentage)
13
Demonstration of the Levels at which the Coefficients should be to Honour Real ‘LTFR’,
Developing Country (Percentage)
13
Demonstration of the Levels at which the Coefficients should be to Honour Technical
‘LTFR’, Developed Country (Percentage)
13
Demonstration of the Levels at which the Coefficients should be to Honour Technical
‘LTFR’, Developing Country (Percentage)
13
Hypothetical Application of the Implementation Period for a Developing Country Assuming
Hypothetical Bound Tariff Figures in Percentage
18
Electronics & Electrical Goods
20
Fish and Fish Products
20
Footwear
20
Leather Goods
21
Motor Vehicles Parts and Equipments
21
Stones, Gems and Precious Metals
21
Textiles and Clothing
21
Electronics and Electrical Goods Sector
22
Fish and Fish Products
22
Footwear
22
Leather Goods
23
Motor Vehicles Parts and Components
23
Stones, Gems and Precious Metals
23
Textiles and Clothing
23
Electronics & Electrical Goods
24
Fish and Fish Products
24
Footwear
24
Leather Goods
25
Motor Vehicles Parts and Equipments
25
Stones, Gems and Precious Stones
25
Textiles and Clothing
25
List ofBoxes
Box 1 Paragraph 14 of the Hong Kong Ministerial Declaration
Box 2 Limitation of the Pakistani Proposal in Light of Paragraph 14 of the Hong Kong
5
Ministerial Declaration
Box 3 Mark up for Countries whose Average Bound Tariff Rate is Low
15
16
yj
Industrial Tariffs and South Asia~
Executive Summary
Industrial tariffs have played a significant role
in the industrial and economic development of
many countries. The present developed countries
allowed their industries to develop behind high
tariff walls. Hence, it is imperative that developing
countries of today should also have the option of
using industrial tariffs as a judicious policy tool to
develop their industrial base. Further, high tariffs
in developed countries especially on products of
export interest to developing countries deny market
access opportunities to developing countries.
Hence, negotiations on industrial tariffs should
address two crucial issues for developing countries.
First, the negotiations should not lead to any drastic
reduction in the industrial tariffs of these countries.
If there is a drastic reduction, it will impose harsh
adjustment costs on developing countries and may
have an adverse impact. Secondly, the negotiations
should result in high tariffs in developed countries,
especially on products of export interest to
developing countries, to reduce or eliminate
drastically.
This was also the objective of the Doha round
of negotiations held in Doha in 2001. The
negotiations on industrial tariffs or Non-Agriculture
Market Access (NAMA) have witnessed a see-saw
effect. Several deadlines have been missed. The
latest deadline, which was agreed at the Hong
Kong ministerial conference, whereby modalities
were to be established by 30 April 2006 has also
been missed. However, what is worrying is not a
delay in finalizing the modalities or the 30 April
deadline being missed but the trend and direction
of negotiations especially on the issue of tariff
reduction in NAMA. The negotiations, on how
to cut the industrial tariffs has become very one
dimensional with developing countries like India
and Brazil abdicating their original stand. I hey are
slowly but surely getting rapt in the negotiating
trap of developed countries.
It is important to understand how this is happening.
Developing countries like India and Brazil have
always been against the pure ‘Swiss formula’ that
developed countries have been advocating, to
cut industrial tariffs primarily because the Swiss
formula cuts tariffs steeply with higher tariffs
being cut even more sharply. Such stringent tariff
reductions will attack the available policy space of
developing countries. To counter such a formula,
India and Brazil along with Argentina, in 2005,
proposed a modified version of the Swiss formula
that takes into account the average tariff rate of a
country while cutting a particular tariff rate. This
formula came to be popularly known as the ABI
formula. Although it met with stiff resistance
from developed countries, India and Brazil were
successful in getting it on the negotiating table in
the Hong Kong (HK) ministerial conference of
the WTO. HK witnessed the adoption of a 'Swiss
formula with coefficients’ for cutting industrial
tariffs. The word ‘coefficients’ in plural meant that
the ABI formula is alive.
A “Swiss formula with coefficients’ implies two
things. First, there could be a simple Swiss formula
that has two coefficients; one for developed and
the other for developing countries. For instance, a
simple Swiss formula with the single coefficient in
the formula being five for developed countries and
15 for developing countries. This is a simple Swiss
formula, as it has only one coefficient at a time
when applied to a particular tariff line. Secondly,
there could be another Swiss formula, which has
one, two, three or n number of coefficients with all
of them a part of the formula. For example, there
may be a Swiss formula, which says that apart from
the single coefficient, say five, the country’s average
tariff rate, say X, will also be a part of the formula.
In this case we have a formula that comprises of
two coefficients at a time when it is being applied
to cut a particular tariff rate.
Now, what needs to be decided in Geneva is which
of these two interpretations should be adopted.
India, Argentina, Brazil and the Caribbean countries
will certainly like the second interpretation to be
adopted. In other words, they will prefer to have
a Swiss formula where factors that reflect the tariff
structures of individual countries such as rhe
average tariff rate or the effect of reduction of tariffs
on revenue are taken into account while reducing
industrial tariffs.
However, the post Hong Kong trends clearly
demonstrate that Brazil and perhaps India, are
negotiating for the first interpretation of ‘Swiss
formula with coefficients’ and not pushing for the
AB1 approach that they had proposed in 2005.
Post Hong Kong, the position of India is not very
clear on the issue of tariff reduction in NAMA. It is
Industrial Tariffs and South Asia
uncertain whether India plans to go ahead with the
first interpretation or still wants to stick to the ABI
proposal. Since Brazil and India have till now been
negotiating, as a group on this issue, there are reasons
to believe that India may also go the Brazil way. If the
Swiss formula with two coefficients is accepted and
assuming that the coefficient of developing country
is 30, then, India will have to make stringent tariff
reductions. For instance, with ‘30’ as the coefficient,
the bound tariff rate offish and fish products in India
will come down from the present 100.7 percent to
23 percent, which is a reduction of as high as 77
percent. This new bound tariff rate would be even
lower than the present applied rate of 30 percent on
fish and fish products in India and hence, complete
erosion of policy space.
This paper argues that the ABI formula is the only
way to go ahead, as it will alone fulfill the mandate
of Less Than Full Reciprocity (LTFR) in reduction
commitments and lead to elimination of tariff
escalation and tariff peak in developed countries.
This paper has used the tariff profile of India and
Pakistan to build a case for the adoption of the ABI
formula as the tariff reduction formula in NAMA.
1. Introduction
Industrial tariffs have historically been recognised
as important policy tools for countries to foster
industrialisation and to follow a host ofdevelopment
friendly domestic policies. There is ample economic
literature to show that developed countries of
today and many other developing countries used
industrial tariffs to protect and promote their
infant industries and develop their industrial bases.
For instance, the United States (US), for rhe most
part of the period from 1820 to 1945, maintained
an average industrial tariff rate of 40 percent and
never less than 25 percent except for brief periods.
Similarly, the East Asian success story also reveals
that protection in the form of higher tariffs has
enabled these countries to develop their industries
and economies. Given this historical importance
and role that industrial tariffs have performed in
the growth of industrial sectors in developed world,
developing countries of today should make full use
of industrial tariffs as an important policy tool.
Notwithstanding, this importance of industrial
tariffs, it is frequently argued by the advocates of
free trade that customs duties and other tariff rates
often act as impediments to free flow of goods across
borders and hence should be substantially reduced.
Article XXVIII bis of the General Agreement
on Tariffs and Trade (GATT) 1994 reflects this
principle of international trade. However, Article
XXVIII bis also states that negotiations directed
towards substantial reduction of the general level
of tariffs and other charges will be done on a
mutually advantageous basis. Further paragraph
3 of Article XXVIII bis of GATT 1994 states
that while conducting negotiations on bringing
down the tariff rates it is important to take due
cognisance of the needs of individual countries
which includes fiscal, developmental, strategic and
other needs. Thus, there is a legal requirement that
any negotiation process or final agreement on tariff
reduction should be consistent with the principles
laid down in Article XXVIII bis of GATT In other
words, tariff reduction should not be in a manner
that takes away the sovereign right of countries to
use tariffs as a policy tool to foster industrialisation
or other developmental policies that it may deem fit.
If tariffs are to be reduced, they should be reduced
gradually and in line with the growing capabilities
of countries that are undergoing liberalisation.1
In the name of liberalising international trade
counties have been engaged with each other,
right from 1947 - the formation of GATT — to
date to reduce industrial tariffs. Successive rounds
of negotiations upto the formation of the World
Trade Organisation (WTO) led to substantial
reduction in industrial tariffs worldwide. After
the formation of the WTO in 1995, yet another
attempt to reduce industrial tariffs was flagged at
the Doha Ministerial conference in 2001 as a part
of the Doha work programme.
However, the present attempt to reduce industrial
tariffs as part of the Doha work programme is
qualitatively different from the earlier rounds of
industrial tariff reduction in terms of its impact on
the industry and economy of developing countries.
Let us try to understand this briefly. Tariffs have
always been one of the tools available to countries
to protect their domestic industry apart from direct
import control measures like quantitative restrictions
(QR). However, post WTO, tariffs have become the
1 For more on this see Ha-Joon Chang, ‘Why Developing Countries Need Tariffs’, Oxfam International and South Centre (2005).
Interpreting for Development
1
only tool available to protect domestic industry as
the QR regime has been dismantled for most of
the countries including India and the balance of
payment measures under Article XVIII B of GAIT
have been put under severe disciplines.2 Hence,
the biggest apprehension that developing countries
like India have is that any hasty and indiscriminate
tariff liberalisation will not only take away the only
available protection and impose harsh adjustment
costs but also devoid the opportunity to use industrial
tariffs to develop industrial base.
These harsh adjustment costs could be in the form of
balance of payment problems, de-industrialisation
and hence unemployment. The Sub Saharan Africa
experience where the rapid reduction in industrial
tariffs durins; 1980s led to de-industrialisation and
unemployment lends weight to these concerns.
Moreover, significant market opening could also
lead to lower tax revenue and hamper interests of
many developing countries as these countries rely
on import tariff for revenue generation.
However, negotiations on industrial tariffs are not
just about developing countries and LDCs adopting
defensive postures but also about the rife double
standards and protectionist stance of developed
countries. On the one hand developed countries apply
enormous pressure on developing world to reduce
tariffs but on the other they themselves maintain
high tariffs especially on products of export interest
to developing countries such as textiles, clothing,
leather, footwear etc. The potential gains that could
accrue to developingcoun tries and LDCs ifdeveloped
countries reduce their high tariff rates on products of
export interest to these countries is mammoth. The
textile, clothing, leather and footwear sector in many
developing countries are labour intensive. If assured
export markets are available for these products it will
not only boost employment but also make a dent
on poverty. Keeping the above-mentioned points in
mind negotiations on NAMA should allow:
Developing countries to retain the policy space
to use industrial tariffs as effective tools to
pursue relevant developmental goals.
2. Developed countries to eliminate or reduce
tariff peaks and tariff escalation to give better
market access to developed countries.
For retention of policy space it is important for
developing counties such as India to retain enough
‘water’ (difference between bound and applied) in
the tariff lines. If there is enough or adequate ‘water’,
then developing countries have the flexibility to
increase their applied tariff rates if the situation
so demands. For instance, India has a bound tariff
rate of 34.3 percent and an applied tariff rate of
27.7 percent. The existing ‘water’ gives India the
flexibility to increase its tariff rate from 27.7 percent
to 34.3 percent if the situation so demands. This
flexibility is extremely important to pursue various
developmental goals that a sovereign country may
choose to pursue. Hence, a tariff reduction formula
should be such that does not take away the ‘water’
in the tariff lines, at least not completely. The
discussion in this paper would focus on this core
issue. It would attempt to argue for an industrial
tariff reduction formula that allows South Asian
countries to have enough ‘water’ in their tariff
structure.
1.
The South Asian region comprising of three
developing countries and two LDCs is an
important illustration of the potential downsides
of an imbalanced NAMA outcome.3 All of these
countries except Sri Lanka to some extent,
maintain high industrial tariff rates. The tariff
profiles of these countries are characterised by
low binding coverage and high international
peaks (See Table 1). Countries such as India have
about 60.1 percent of tariff lines that are more
than 15 percent (International Peaks). Hence,
the whole of South Asia has a high industrial
tariff profile.
2 Bhagirath Lal Das, ‘Market Access for Non-Agriculrural Products’ in ‘The New Work Programme of the WTO , Third World
Network (2002).
’ South Asia in this context refers to Bangladesh, India, Nepal, Pakistan and Sri Lanka.
2
Industrial Tariffs and South Asia
Given such a high tariff profile and from the
experience of Sub Saharan Africa it is amply clear
that any significant or drastic opening up of these
markets may not augur well for the South Asian
countries. At the same time these countries face a
lot of barriers in markets of developed countries in
terms of tariff peaks and tariff escalation. Tariff peaks
or high tariffs continue to be maintained in sectors
of export interest to developing countries. Further,
tariff escalation as a tariff measure in developed
countries encourages imports of raw materials
and discourages the imports of value added goods
and thereby of processing industry in developing
countries. It does not allow developing countries
and LDCs to graduate from exporting raw materials
to processed and finished goods. J he average US
tariff for all imports is 1.6 percent, but this rises
to 4 percent for imports from India and to 14-15
percent for imports from LDCs such as Bangladesh
and Nepal. EC imposes tariffs of less than 4 percent
on Indian yarns, but this tariff rate escalates to 12
percent if the yarn is woven into garments.
TABLE 1
Industrial Tariff Profile of South Asian Countries
Binding Coverage4
Simple Average
International
Peaks5
National Peaks6
Bangladesh
3.1
42.9
2.7
0.2
India
69.8
34.3
60.1
0.1
Pakistan
36.9
35.3
33.2
0.0
Sri Lanka
28.3
19.2
13.1
0.5
Import Market
Source: World Trade Organisation Secretariat, WTO Member’s Tariff Profiles, TN/MA/S/4/Rev. 1/Corr. 1.15 November 2002
2. NAMA Negotiations
The elements of Article XXVIII bis of GATT 1994
were reflected in the Doha Development Agenda7
(DDA). Paragraph 16 of the DDA states that
negotiations on NAMA will in future target to reduce
or eliminate all kinds of tariffbarriers in particular on
products of export interest to developing countries.
It also states that negotiations shall take cognisance
of the development needs of developing and least
developed countries. It also recognises the special
and differential treatment for developing countries
by asking them to make less than full reciprocity
commitments. In other words, developing and
least developed countries should not be asked to
undertake the same reduction commitments that
developed countries undertake.8
After the adoption of the DDA, negotiations
on non-agricultural products were launched in
January 2002 by creating a negotiating- group
on market access (NGMA). The ongoing
negotiations on non-agricultural products were to
be completed by 1 January 2005. However, this
deadline has been missed and the negotiations are
ongoing with no firm end dates in sight.
The participants in the negotiation process were
first expected to agree on how to conduct the tariff
cutting exercise. In other words the participants
first have to agree on ‘modalities’. The current
state of play is that all the member countries
are still struggling to establish modalities for
4 Binding Coverage implies the extent of tariff lines that have bound tariff rates.
5 International Peaks shows the percentage of tariff lines in a country that have a bound tariff rate of more than 15 percent.
6 National Peaks shows the percentage of tariff lines in a country that have bound tariff rates at least three times higher than the country's simple
average.
7 World Trade Organisation Ministerial Declaration, WT/MIN (01 J/DEC/l, adopted 14 November 2001, para 16.
8 Prabhash Ranjan (2005), Tariff Negotiations in NAMA and South Asia: July Agreement and Beyond, Working Paper 3, Centre for Trade and
Development.
Interpreting for Development
3
future negotiations, though, in July 2004 all the
member countries had agreed to a framework for
establishing modalities in market access for nonagricultural products.9 The July agreement was an
important development as it was the first agreement
amongst the member countries of the WTO after
the collapse of the Cancun ministerial. Also, since
it lays down the framework for establishing future
modalities it needs to be comprehended well as it
determines the future course of action. After the
July agreement in 2004, a lot of water has flown.
Intensive negotiations resulted in some sort of
agreement on NAMA emerging during the Vlth
ministerial conference of the WTO at Hong Kono
(HK). The HK ministerial conference saw the
emergence of the following important issues:1011
1. A 'Swiss formula with coefficients’ will be
adopted for undertaking tariff reduction.
2. Reiteration of flexibilities to developing
countries
3. Participation in the sectorals will be voluntary
4. A non-linear mark up will be used to bind the
unbound tariff lines.
3. Scope of the Paper
This paper does not look at all the issues pertaining
to NAMA negotiations. Instead, it analyses one of
the most contentious issue in NAMA negotiations
i.e. the tariff reduction modality. It makes an attempt
to examine the HK declaration on the core issue of
tariff reduction for industrial goods. It also looks at
how the proposed architecture of the tariff reduction
formula will affect the tariff profiles of South Asian
countries. Building on the potential impact of the
different tariff reduction formula on South Asian
countries, this paper would argue fora tariff reduction
formula which can produce a relatively balanced
NAMA outcome. This argument also stems from
the fact that for South Asian countries to maintain
‘water’ in their tariff structure, it is important to
clinch an appropriate tariff reduction formula. In
case of an agreement on a formula, only India and
Pakistan within South Asia will have to undertake
tariff reduction". Bangladesh and Nepal do not
need to cut tariffs because of their LDC status and
Sri Lanka is also exempted since its binding coverage
is less than 35 percent. The paper would also develop
an interpretation of the possible architecture of a
tariff reduction formula supportive of the interests
of developing countries in general and South Asia
in particular.
4. General Comment on NAMA Outcome of HK
It is important to look at the history of the NAMA
negotiations in order to understand the context
of the present agreement on NAMA. Developing
countries had vociferously opposed the NAMA
portion of the Draft Cancun ministerial text. The
reason behind this opposition was that the draft text
of Cancun talked of having a non-linear formula for
tariff reduction, limited flexibilities and an unviable
clause for the treatment of the unbound tariff lines.
At Cancun, developing counties succeeded in
blocking the draft text.
9 Decision Adopted by the General Council, WT/L/59, adopted 1 August 2004, Annex B
10 Draft Ministerial Declaration, WT/MIN (05)/W/3/Rev.2, 18 December 2005
11 The impact of the tariff rates of India and Pakistan in this paper is shown at HS 6 digit and at HS 8 digit level respectively. For Pakistan the
analysis is at HS 8 digit level for lack of data at the HS 6 level. However, this will not affect the overall argument.
4
Industrial Tariffs and South Asia
Developing countries expected that this draft text
on NAMA would not come up again in the future
negotiations. However, to their surprise, they saw
the same text coming up in the negotiations at
the General Council in July 2004. The first draft
of the July framework agreement that came up on
17 July 2004 had the exactly same NAMA text
that developing countries had rejected at Cancun.
Developing countries opposed this again and
mainly due to rhe efforts of the African countries
an additional paragraph was added to this
controversial text. This was the first paragraph in
Annex B of the July framework agreement, which
said that whatever is given below is not final. This
paragraph gave a window to developing countries
to negotiate for a linear formula for tariff reduction
amongst other development friendly positions.
However, developing countries did not use this
window at all. In the build up rhe Hong Kong
ministerial, they negotiated as if this additional
paragraph did not exist. In other words, there was
a tacit acceptance of the text that they themselves
had opposed and rejected. It can then be said
that at Hong Kong the biggest casualty was the
cementing of Annex B of the July framework
minus the first paragraph which should have
been at the centre of negotiations. So, that
which was inconclusive and negotiable has now
become an inseparable part of the multilateral
negotiations.
5. Swiss Formula with Coefficients
One of the few things that were agreed by countries
in NAMA was the adoption ofa 'Swiss formula with
coefficients’ for cutting industrial tariffs (Paragraph
14 of the Hong Kong Declaration) (See Box 1).
However, a ‘Swiss formula with coefficients’ has
created a lot of confusion. Many have interpreted
this as the acceptance of the highly iniquitous 'Swiss
formula’ that cuts industrial tariff rates steeply and
in fact cuts higher tariff rates even more steeply.
However, this is not true and part of the reason for
this kind of interpretation doing rounds is that the
literature on industrial tariffs in the WTO has often
demarcated between a 'Swiss formula and a 'Swiss
type or a modified Swiss formula’. Therefore, each
time we see the words 'Swiss formula’ it makes us
think that the bad and ugly ‘Swiss formula’ is back
again and has now embedded in the multilateral
trade negotiations.
BOX 1
Paragraph 14 of the Hong Kong Ministerial
Declaration
Adopt a Swiss Formula with coefficients at levels
which shall inter alia:
•
•
Reduce or as appropriate eliminate tariffs,
including the reduction or elimination of
tariff peaks, high tariffs and tariff escalation,
in particular on products of export interest
to developing countries.
Take fully into account the special needs
and interests of developing countries,
including through less than full reciprocity'
in reduction commitments.
We instruct the Negotiating Group to finalize its
structure and details as soon as possible.
A ‘modified Swiss formula’ could be written as:
At the conceptual level it is important to understand
the distinction between a 'Swiss formula’ and
a 'modified Swiss formula’, which is also a Swiss
formula. Swiss formula is written as:
RyT
T\ =——— , where T1 is the final tariff rate,
T is the initial tariff rate and B is a coefficient.
71 =
x
x
(B *X) + T
Where T1 is the final tariff rate,
T is the initial tariff rate, B and X are coefficients.
The difference between the 'modified Swiss
formula’ and the 'Swiss formula’ is the presence
of the additional coefficient X. It is important to
Interpreting for Development
5
understand that this coefficient is only an additional
feature in the ‘Swiss formula’ mentioned above and
has an impact on the existing coefficient B. So
(B x X) is only a new coefficient that we have in
the normally understood ‘Swiss formula’. This
new coefficient will be either greater than B or
smaller than it depending on the value of X. So, if
B x X = Y, then the so called ‘modified Swiss
formula’ will become a ‘Swiss formula’ and can be
represented as:
yxT
71 =-------- , where Y is a coefficient and T1 and
T+T
T are the final and initial tariff rates. So, we find
that this so called ‘modified Swiss formula’ is also
essentially a ‘Swiss formula’. The only difference is
rhat the so-called ‘modified Swiss formula’ has one
more factor (X in the example above) or coefficient
embedded in the formula.
5.1 Possible Interpretations of
Paragraph 14
After this conceptual understanding, it is important
to understand how the countries will negotiate
towards establishing modalities for a tariff reduction
formula. The adoption of a ‘Swiss formula with
coefficients’ implies two things. Firstly, that a Swiss
formula will be adopted and secondly, that it will
have more than one coefficient. Reading both these
provisions together implies that a Swiss formula,
which has only one coefficient for both developed
and developing countries, will not be adopted.
The word ‘coefficients’ in paragraph 14 means two
things:
1. A Swiss formula, with two different
coefficients, one each for developed and
developing countries. Such a formula could be
represented as:
RvT
T'l =-------- , where T1 is the final tariff rate, T is
B+ T
the initial tariff rate and B is a coefficient which
has different values for developed and developing
countries. B = m for developed countries, and
B = n for developing countries.
6_
Industrial Tariffs and South Asia
This formula is a Simple Swiss formula and is similar
to one that has been proposed by countries like the
US or even Pakistan. The difference in the proposal
of US and Pakistan is the value of the coefficient
for developed and developing countries.
2. A Swiss formula, with 2, 3, or n number
of coefficients with these coefficients being
embedded in the formula unlike the formula
given above, which has more than one coefficient
and hence is a ‘Swiss formula with coefficients’
but has only one coefficient embedded in the
formula.
This kind of formula could be represented in many
ways. Below are the two most important ways in
which such a formula could be represented:
T'l =
x
x
(BxX)+T
X is the average tariff rate of a
&
country. This has been discussed above. This type
of formula has been proposed by Argentina, Brazil
and India.
Another manner in which this formula could be
represented is as follows:
{(B + C)xX}*T
,
. , _ .
J 1 =-------------------------- , where 1 1 is the hnal tariff
{(B + C) x Xi + T
rate, T is the initial tariff rate, B is a coefficient, X
is the average tariff of a country and C is the credit
to be accorded to a developing country.
One may argue that rhe formula rhat has been
proposed by the Caribbean group of countries is not
a Swiss formula within the mandate of the Hong
Kong Declaration. This is not true. {(B+C)} x X} x F
will also yield a single coefficient. Assuming that such
a coefficient is, K and replacing ((B+C)|xX|x f
by K in the above formula we get:
jy-
'r
74 =___ — , which is a Swiss formula.
K+ T
The question that arises is which of these two
interpretations shall be adopted. Before, we come
to the issue of which of these two interpretations
should be adopted, it is important to see the impact
that each of these formulas will have on the present
bound tariff rates of India and Pakistan.
developing country. 1 he argument behind having
low coefficients such as 15 or 20 for developing
countries is to have deeper cuts in the existing
tariffs. EU and US have been arguing that the
Impact of the Possible Tariff Reduction Formulas
on Tariff lines of India and Pakistan
use of these coefficients will result in bringing
the existing bound tariffs in countries such as
India below the existing applied tariff rates and
hence achieve real market access. However, on
the other hand, countries like India have argued
that they have already given real market access by
autonomously lowering their tariff rates.
5.2 Swiss Formula with Two Coefficients
The first possible tariff reduction formula could be
the Swiss formula with two coefficients.
„
R xT
T1 = -—— , where B = 5, 10, 15, 20, 30, 35
B +T
Let us now look at some simulations for India
and Pakistan for different coefficients. From Table
2, it is clear that lesser the value of the coefficient
steeper is the reduction. For instance, if a coefficient
of‘15’ is used, then the bound tariff rate of 100
percent will come down to 13 percent, which is
approximately a reduction of 87 percent. This is a
very steep reduction and will bring down the new
bound tariff rate much below the present applied
tariff rate. Similarly, in case of Pakistan, the same
After the HK ministerial conference, several
simulations have been done to see the impact of
different coefficients on the existing tariff rates
of many developing countries. However, in all
these simulations, the first option of having two
coefficients; one for developed and the other
for developing have been used. Some of these
coefficients are: 10 & 15, 8 Sc 20 etc. The first
coefficient here represents the coefficient for
developed country and the other coefficient for
TABLE 2 A
Impact of the Pure Swiss Formula on the Bound Tariff Rates of 5107 Tariff Lines of India (HS 6
Digit Level) (Percentage)
Initial Tariff
Rate (TO)
No. of Tariff
Lines
0
53
Final Tariff Rate (Tl)
B=10
B = 15
B=20
B=30
B=35
0
0
0
0
0
0
B=5
3
7
1.88
2.31
2.50
2.61
2.73
2.76
5
76
2.50
3.33
3.75
4.00
4.29
4.38
10
25
3.33
5.00
6.00
6.67
7.50
7.78
15
147
3.75
6.00
7.50
8.57
10.00
10.50
20
150
4.00
6.67
8.57
10.00
12.00
12.73
25
954
4.17
7.14
9.38
11.11
13.64
14.58
30
73
4.29
7.78
10.00
12.00
15.00
16.15
35
3512
4.38
7.78
10.50
12.73
16.15
17.50
40
76
4.44
8.00
10.91
13.33
17.14
18.67
45
6
4.50
8.18
11.25
13.85
18.00
19.69
50
4
4.55
8.33
11.54
14.29
18.75
20.59
55
1
4.58
8.46
11.79
14.67
19.41
21.39
60
7
4.62
8.57
12.00
15.00
20.00
22.11
(Contd..... )
Interpreting for Development
7
(Table Contd.... )
Initial Tariff
Rate (TO)
No. of Tariff
Lines
Final Tariff Rate (Tl)
70
15
4.67
8.75
12.35
75
2
4.69
8.82
80
4
4.71
8.89
100
38
4.76
105
15
4.77
15.56
21.00
23.33
12.50
15.79
21.43
23.86
12.63
16.00
21.82
24.35
9.09
13.04
16.67
23.08
25.93
9.13
13.13
16.80
23.33
26.25
115
1
4.79
9.20
13.27
17.04
23.79
26.83
170
2
4.86
9.44
13.78
17.89
25.50
29.02
8
4.88
9.55
14.00
18.26
26.25
30.00
210
Total = 5107
Source: Authors Calculations from UNCTAD figures
TABLE 2B
Impact of the Pure Swiss Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistan
(HS 8 Digit Level) (Percentage)
-
Final Tariff Rate (Tl)
Initial Tariff
Rate (TO)
No. of Tariff Lines
B=5
B = 10
20
21
4.00
30
67
4.29
40
122
50
838
B=15
B=20
B=30
6.67
8.57
10.00
12.00
12.73
7.50
10.00
12.00
15.00
16.15
4.44
8.00
10.91
13.33
17.14
18.67
4.55
8.33
11.54
14.29
18.75
20.59
B=35
Total = 1048
Source: Authors Calculations from WTO figures
coefficient will bring down the bound tariff rate
of 50 percent to about 11.5 percent, which is a
reduction of about as high as 77 percent.
5.3 ABI Formula
The tariff reduction formula proposed by
Argentina, Brazil and India is called the ABI
formula. This formula talks of having the
average tariff rate of a country as one of the
coefficient in the formula. The presence of the
average tariff rate of a country as the coefficient
will ensure that country’s present tariff profile
is taken into account while undertaking tariff
reduction.
3
Industrial Tariffs and South Asia
x T , where X is the average tariff
(B xX) + T
rate of a country. X = 34.3 percent for India (does
not include specific duties) and 35.3 percent for
Pakistan (does not include specific duties).
T\ =
The application of the ABI formula on the
tariff lines of India and Pakistan demonstrates
that this formula does not lead to a very steep
reduction. However, the reduction in this case
can also be steep if the value of B is less than
1. As, the value of B increases the reduction is
less steep. For instance, in the case of India, the
bound tariff rate of 100 percent will come down
to 25.54 percent (when B = 1), as against the
TABLE 3A
Impact of the ABI Formula on the Bound Tariff Rates of 5107 Tariff Lines of Indian Industrial
Products at HS 6 Digit Level (Percentage), the Average Tariff Rate = 34.3 Percent (Percentage)
Initial Tariff
Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
i
B = 0.5
B=1
B = 1.5
B=2
B = 1.75
0
53
0
0
0
0
0
3
7
2.55
2.76
2.83
2.86
2.87
5
76
3.87
4.36
4.56
4.62
4.66
10
25
6.32
7.74
8.37
8.57
8.73
15
147
8.00
10.44
11.61
12.00
12.31
20
150
9.23
12.63
14.40
15.00
15.49
25
954
10.17
14.46
16.82
17.65
18.32
30
73
11.51
17.32
20.83
22.11
23.18
35
3512
11.51
17.32
20.83
22.11
23.18
40
7
12.00
18.47
22.50
24.00
25.17
45
6
12.42
19.46
24.00
25.72
27.17
50
4
12.77
20.34
25.36
27.28
28.92
55
1
13.07
21.13
26.58
28.70
30.53
60
7
13.34
21.82
27.70
30.01
32.01
70
15
13.78
23.02
29.65
32.31
34.65
75
2
13.96
23.54
30.52
33.34
35.83
80
4
14.12
24.01
31.31
34.29
36.93
100
38
14.64
25.54
33.97
37.51
40.69
105
15
14.74
25.85
34.53
38.19
41.49
115
1
14.92
26.42
35.55
39.44
42.97
170
2
15.58
28.54
39.50
44.36
48.88
8
15.86
29.48
41.33
46.68
51.71
210
Total = 5107
Source: Author's Calculations from WTO figures
tariff rate of 13 percent, when the coefficient
of ‘B’ is 1 5.
cutting exercise will be fair and the competing needs
of different countries can be accommodated.
5.4 Caribbean Formula
T\ = - --------------------- '------ , where T1 is the final tariff
This formula was proposed by the Caribbean
countries. The rationale for this formula is that
countries use tariffs for numerous purposes such as
generating revenue and therefore these factors should
also be taken into account while cutting tariff rates.
If such factors are taken into account then the tariff
rate, T is the initial tariff rate, B = 112 (a coefficient),
X is the average tariff of a country (34.3 percent
for India and 35.3 percent for Pakistan) and C =
1, 2, 3, 4 (credit to be accorded to a developing
country).
+ C) xA] xT
{(B + C) xX| +T
i
-r, •
,
c
,
-rr
12 The value of B could be anything. However, for the sake of simplicity, it has been assumed that the value of‘B’ here is 1.
Interpreting for Development
9
TABLE 3B
Impact of the ABI Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistani Industrial
Products at HS 8 Digit Level (Percentage), the Average Tariff Rate = 35.3 Percent (Percentage)
1 Initial Tariff Rate (TO)
No. of
Tariff Lines
Final Tariff Rate (Tl)
B=0.5
B=1
B=1.5
B=2
20
21
9.38
6.67
14.52
15.58
30
67
11.11
16.22
19.15
21.05
40
122
12.25
18.75
22.79
25.53
50
838
13.09
20.69
25.72
29.27
Total = 1048
Source: Authors Calculations front WTO figures
TABLE 4A
Impact of the Caribbean Formula on the Bound Tariff Rates of 5107 Tariff Lines of Indian
Industrial Products at HS 6 Digit Level (Percentage), the Average Tariff Rate = 34.3 Percent,
B = 1 (Percentage)
Final Tariff Rate (Tl)
Initial Tariff Rate (TO)
No. of
Tariff Lines
C =1
C =2
C =3
C =4
0
53
0
0
0
0
3
7
2.87
2.92
2.94
2.95
5
76
4.66
4.77
4.82
4.86
10
25
8.73
9.11
9.32
9.45
13.79
15
147
12.31
13.09
13.52
20
150
15.49
16.75
17.46
17.91
25
954
18.32
20.11
21.15
21.82
30
73
20.87
23.23
24.62
25.53
35
3512
23.18
26.12
27.89
29.07
40
7
25.27
28.80
30.97
32.43
45
6
27.17
31.31
33.89
35.65
36.65
38.71
50
4
28.92
33.65
55
1
30.53
35.84
39.26
41.64
44.45
60
7
32.01
37.90
41.74
70
15
34.65
41.66
46.35
49.71
75
2
35.83
43.38
48.49
52.18
80
4
36.93
45.01
50.53
54.55
100
38
40.69
50.71
57.84
63.17
65.13
105
15
41.49
51.97
59.48
115
1
42.97
54.31
62.56
68.84
48.88
64.10
75.92
85.37
51.71
69.06
82.98
94.40
2
170
8
210
Total = 5107
Source: Author’s Calculations from UNCTAD figures
10
Industrial Tariffs and South Asia
TABLE 4B
Impact of the Caribbean Formula on the Bound Tariff Rates of 1048 Tariff Lines of Pakistani
Industrial Products at HS 8 Digit Level (Percentage), the Average Tariff Rate = 35.3 Percent,
B = 1 (Percentage)
Initial Tariff Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
C= 1
C= 2
C=3
C=4
21
15.58
16.82
17.52
17.96
30
67
21.05
23.38
24.74
25.64
40
122
25.53
29.03
31.17
32.61
838
29.27
33.96
36.92
38.96
20
50
Total = 1048
Source: Author’s Calculations from WTO figures
The credit to be accorded will depend on factors
such as tariff revenue and other role that tariffs
are playing within a country. The greater the role
more will be the credit that will be accorded to a
particular country. In the simulations for India
and Pakistan the impact of the Caribbean formula
has been demonstrated byJ assigning
o o different
credits. One can see that as the value of credit
assigned increases the tariff rate reduction becomes
less steep.
Which Interpretation to Accept?
From the impact of the three possible formulas
on the tariff rates of India and Pakistan one can
safely argue that the Caribbean formula and the
ABI formula are much less stringent as compared
to the ‘Swiss formula with two coefficients’.
Therefore, from the perspective of developing
countries including those of South Asia, ‘Swiss
formula with coefficients’ should be interpreted as
a Swiss formula that has more than one coefficient
embedded in the formula.
Let’s see how best the HK declaration on reducing
the industrial tariffs can be interpreted to
accommodate the above findings. In other words,
how best can paragraph 14 be interpreted to build
the case for having the ABI or the Caribbean
formula as against the ‘Swiss formula with two
coefficients’.
Paragraph 14, apart from stating that a ‘Swiss
formula with coefficients’ will be adopted, also
states that the coefficients be adopted at levels
which ‘shall’
1. take fully into account the special needs of
developing countries including through LTFR
in reduction commitments.
2. reduce tariff peaks and tariff escalation
In other words, the declaration makes it binding
and mandatory (as implied by the word ‘shall’)
to have such coefficients in the formula that will
eliminate tariff peaks and tariff escalation and also
honour less than full reciprocity (LTFR). A set of
coefficients that do not fulfil these two requirements
cannot become a part of the tariff reduction
formula. In this entire equation, the interpretation
of‘LTFR’ holds the key to the determination of the
coefficients in the Swiss formula.
5.5 Less Than Full Reciprocity (LTFR)
The interpretation of LTFR has always been a moot
issue. Different countries have attempted different
interpretations to suit their interest. However, the
most common and acceptable interpretation is
that developing countries should undertake less
stringent obligations than what developed countries
will undertake.
Interpreting for Development
11
This can be understood with the help of the
following example. Imagine that the initial tariff
rate of a tariff line ‘A’ is 10 percent in a developed
country ‘M’ and the initial tariff rate of the same
tariff line in a developing country ‘N’ is 50 percent.
Now if ‘M’ agrees to reduce the tariff rare by 70
percent, then ‘N’ should reduce its tariff rate bv
any value that is less than 70 percent and not by
70 percent or more. So, in other words, honouring
‘LTFR’ in context of industrial tariffs implies that
developing countries undertake lesser cuts than
what developed countries will undertake. In this
example let’s assume that developing country ‘N’
cuts its tariff rate by 65 percent (See Table 5). The
final tariff rate for this country will be 14.
TABLE 5
Demonstration of LTFR (Percentage)
Initial
Tariff
Reduction
in Initial
Tariff
Final Tariff
■a
Source: Authors calculation.
46 percent (2/3rd of 70). It can cut by less than 46
percent.
However, the principle of real ‘LTFR’ also has its
limitations, especially in cases where the tariff rates
of developed countries are already very low. For
instance, if the tariff rate in a developed country
comes down from 4 percent to zero percent, there
will be a reduction of 100 percent. This reduction
in the tariff rate, however, will change the price only
notionally from 104 units to 100 units (Assuming
100 units is the base price). On the other hand, if a
developing country has to cut its initial tariff of 30
percent by 2/3rd of 100, which is 66 percent, then
the final tariff rate will be 10 percent and the price
of the imported product will come down from 130
units to 110 units, which is a steep reduction in the
price. In this case, although there is real ‘LTFR’,
the developing country stands to lose out, as the
product of the developed country will get cheaper
in the market of the developing country whereas
there will not be any significant change in the price
of the product of the domestic country.
5.6 Implementing Paragraph 14
However, this interpretation of ‘LTFR’ is not
complete as it is a mere technical fulfilment of
‘LTFR’. The other important part is that there
should be a substantial gap between percentage
reduction that developed and developing countries
undertake. Only a substantial gap would ensure
‘LTFR’ in real terms. This brings us to the question
of what is a substantial gap to ensure real ‘LTFR’. In
this regard it can be proposed that for real ‘LTFR’
in industrial tariffs developing countries cannot be
expected to undertake commitments, which would
be more than 2/3rds of commitments made by
developed countries. The reason why a difference
of 2/3rd is being used to operationalise real ‘LTFR’
in industrial tariffs is that the same gap has been
proposed in agriculture. Hence, for real ‘LTFR’ a
developing country should cut its tariff rate by at
least 2/3rd of what a developed country is doing,
or less. In this example, if a developed country
cuts its tariff rate by 70 percent, then a developing
country should not cut its tariff rate by more than
Industrial Tariffs and South Asia
What follows from the above discussion is that the
coefficients in the Swiss formula should be such
that ‘LTFR’ in reduction commitments is really
honoured. On the basis of this we will endeavour
to find out the coefficients by taking the example
of tariff rates of textile and clothing in the US and
India. However, in order to honour LTFR (i.e.
developing country cutting its tariff rates at lesser
rate than a developed country), one needs to first
find out the rate at which developed countries
should cut their tariff rate.
Paragraph 14 is a useful tool to find out the rate at
which developed country should cut its tariff rate.
It states that coefficients in the Swiss formula ‘shall’
eliminate or reduce tariff peaks and tariff escalation.
The bound tariff rate (BTR) of textile and clothing
in the US is 8.6 percent, which is almost three times
the average tariff rate in the US of 3.2 percent for
all non-agricultural products. This is a clear case of
tariff peaks. Hence, a tariff rate as high as 8.6 percent
should come down to at least 3 percent, which
means a reduction of 65 percent. For a reduction of
65 percent the coefficient in the first type of formula
for the US will be 5 (See Table 6A).
Now, as per the interpretation of ‘LTFR’, India
should certainly not cut its tariff rate by more than
2/3rd of 65 percent, which is 43 percent. India’s
BTR for textile and clothing is 26.3 percent.13 After
a reduction of 43 percent, this will come down to
14.2 percent. The coefficient for such a reduction
will be 30 (See Table 6B).
In this case the coefficients in the Swiss formula will
be 5 for developed country and 30 for developing
country. These values of coefficients are completely
different from what developed countries have been
proposing. For instance, the EU proposed that the
value of the coefficient for both developed and
developing countries should be the same. Same
coefficient for developed and developing country
implies violation of‘LTFR’. Similarly, US proposed
that both the coefficients in the Swiss formula
should be ‘within the sight of each other’. In other
words, the coefficients should not be too far away
from each other. One of the fundamental reasons
behind US advocating that the coefficients should
be ‘within the sight of each other’ is to make sure
that bound tariff rates of countries like India come
below the applied tariff rates. This attempt is not
acceptable to India, as it would like to retain ‘policy
space’ by having ‘water’ (difference between bound
and applied tariff rate) in the tariff line.
Besides, this proposal will also not stand the scrutiny
of the HK declaration, as paragraph 14 clearly states
that the coefficients should be at levels, which will
ensure ‘LTFR’. However, if the coefficients are
‘within the sight of each other’, then ‘LTFR’ will
be violated and hence there will be a violation of
paragraph 14. From the above example, it is clear
that for honouring real ‘LTFR’ the two coefficients
have to be at least 6 times apart from each other.
Even if we assume the mere technical fulfilment
of the ‘LTFR principle’ and cut the BTR of textile
and clothing in India by 50 percent, the coefficient
for reduction will be 26, which is still 5 times more
than the coefficient for developed country (See
Table 7A and 7B). This drives home the point that
even for mere technical fulfilment of ‘LTFR’ the
US proposal of having both the coefficients ‘within
the sight of each other’ cannot be accepted.
TABLE 6A
Demonstration of the Levels at which the
Coefficients should be to Honour Real 'LTFR',
Developed Country (Percentage)
Initial
Tariff Rate
Reduction
Final Tariff
Rate
Coefficient
TABLE 6B
Demonstration of the Levels at which the
Coefficients should be to Honour Real 'LTFR',
Developing Country (Percentage)
Initial
Tariff Rate
Reduction
Final Tariff Coefficient
Rate
Source: Authors calculations
TABLE 7A
Demonstration of the Levels at which the
Coefficients should be to Honour Technical
'LTFR', Developed Country (Percentage)
Initial
Tariff Rate
(Tm)
Reduction
in Tm
Final Tariff Coefficient
Rate
for
Developed
Country
TABLE 7B
Demonstration of the Levels at which the
Coefficients should be to Honour Technical
'LTFR', Developing Country (Percentage)
Initial
Tariff Rate
(Tn)
Reduction
in Tn
Final Tariff
Rate
Coefficient
for
Developing
Country
Source: Author's calculations
13 World Trade Report 2004. The BTR for Textile and Clothing (T&C) in India is derived from the ad valorem bound rate of all the tariff
lines in T&C and does not take into account the 271 tariff lines in T&C that have specific duties.
Interpreting for Development
13
So, the advantage of paragraph 14 is that it puts
an end to both the US and the EC proposals of
reducing tariff rates.
The other advantage of paragraph 14 is that it helps
developing countries to make a case for a formula
where the coefficients are far away from each other
so as to honour ‘LTFR’.
5.7 Two Approaches
The levels of coefficients once again bring us to
the two interpretations of the Swiss formula, as
was discussed above. The issue is which of the
two variants is to be used for structuring the
coefficients.
Option 1
The first option is to go for the formula with two
coefficients, one for developed and the other for
developing countries. In such a case the coefficient
could be 5 for developed country and 30 for
developing country.14
So, the formula will be 73 = ——— , where B = 5
B+ T
for the developed country and 30 for the developing
country (See Table 6).
Limitation of Option 1
However, this formula approach has two
limitations. The first limitation is that even with a
coefficient value of 30, there is a steep reduction in
the tariff rate for a country like India. For instance,
if we use 30 as the coefficient to cut the tariff rate
of 35 percent, which is the BTR of majority of
tariff lines in India, then the final tariff rate will
become 16 percent. This is drastic reduction and
will eat up the ‘water’ in the existing tariff lines
and take away the policy space. One of the major
demands of India has been that it wants a tariff
reduction formula that does not completely take
away the existing ‘water’ in the tariff lines.
The second limitation arises when the value of B
is low, say, 2 or even less, in the above formula.
If the value of B is lower, the rate of reduction of
the tariff rate will be even steeper. In such a case
the developed country’s tariff rate will come down
steeply and may cause a steep reduction in the tariff
rate of developing countries as well. For instance, if
the value of B is 2, then 8.6 percent (in the above
example) will come down to 1.6 percent, which is
a reduction of about 81 percent.
Now, if we apply the above-mentioned proposal
of developing countries cutting its tariff rate by
maximum 2/3rd of what developed countries do,
then, India (in this example) will have to reduce
its tariff rate by 54 percent (2/3rd of 81). This
reduction will result in the new tariff rate being 12
percent, which is even less than the applied tariff
rate of textile and clothing. If the final tariff rate is
12 percent, then, the coefficient (B) in the formula
will be 22.
It is interesting to note that although the coefficient
for the in developing country is 11 times more
than the coefficient for the developed country, the
tariff reduction is very steep. Technically, 2 and
22 are coefficients at levels that will honour real
‘LTFR’. However, in reality they will lead to steep
reduction in the tariff rates of developing countries
(India in this case). Hence, this formula option
has its limitation and does not help the cause of
developing countries.
The limitation of this option can also be understood
by taking the US proposal that it has made after
the HK conference, of having two coefficients of
10 & 15. This proposal is similar to what US had
proposed earlier that the two coefficients should be
‘within the sight of each other’. We have already
seen in the analysis above that the proposal to have
the coefficients ‘within the sight of each other’ is in
violation of paragraph 14.
14 This comes close to the proposal made by Pakistan for having a Swiss formula (71 = - “
for developing countries.
*
4
Industrial Tariffs and South Asia
■), where B = 6 for developed countries and 30
BOX 2
Limitation of the Pakistani Proposal in Light of Paragraph 14 of the Hong Kong Ministerial
Declaration
Pakistan, before the HK ministerial conference, had proposed a ‘Swiss formula’
B XT
71 = --------- , where B = 6 for developed countries and 30 for developing countries for tariff
B +T
reduction. This proposal of Pakistan, according to the analysis of this paper, falls in the first type of
interpretation of the ‘Swiss formula with coefficients’. The proposal of Pakistan is of a ‘Swiss formula with
two coefficients’. However, if we apply this formula to the example of Textile and Clothing of US and India
(the example taken in this paper), we will find that it does not honour the principle of LTFR.
With ‘6’ as the coefficient, the BTR ofTextile and Clothing in US will come down from 8.6 percent to 3.5
percent, which is a reduction of 59 percent. With ‘30’ as the coefficient, the BTR ofTextile and Clothing
in India will come down from 26.3 percent to 14 percent, a reduction of 46 percent. However, as per the
definition of LTFR as discussed in this paper, India should reduce its tariff rate by not more than 39 percent
(2/3rd of 59 percent).
Reduction by 39 percent will imply that the new BTR ofTextile and Clothing in India should be 16 percent.
If 16 percent is the new BTR, then the coefficient will be 40 and not 30.
Hence, if the coefficient for developed country is ‘6’ in the ‘Swiss formula with two coefficients’, then the
coefficient for developed country should be ‘40’. Any coefficient less than ‘40’ will result in violation of the
LTFR principle, which according to paragraph 14, has to be honoured by the tariff reduction formula.
This can be demonstrated by taking the example of
the tariff rate ofT&C of US and India. Applying
the Swiss formula with coefficient of 10, the tariff
rate of 8.6 percent will come down to 4.6 percent,
which is a reduction of 46 percent.
coefficient of 62 for developing country would fulfil
the principle of LTFR. From this discussion it is clear
that a ‘Swiss formula with 10 and 15 as the coefficients’
violates paragraph 14 of the HK Declaration and hence
cannot be accepted as the tariff reduction formula.
If we apply the Swiss formula with coefficient of 15
on the tariff rate of T&C of India (26.3 percent),
this tariff rate will come down to 9.5 percent, which
is a reduction of about 63 percent.
The US proposal also states that the tariff rates
should be cut from the applied levels so as to have
the so called ‘real trade flows’. This proposal is not
only anti developmental but also illegal (I hope this
term can be used!). The mandate of the current
negotiations is to cut industrial tariffs from the
bound levels and not applied levels.
Hence, it is clear that if these coefficients are used,
then a country like India will cut its tariff rate of the
same tariff line by much more than what US will do.
On the other hand, ifwe apply the principle of2/3rd,
then India should not cut its tariff rate by more than
30 percent (2/3rd of 46 percent). In such a case the
coefficient will be 62 and not 15. So, if we have the
coefficient of 10 for developed country, then only a
Option 2
The other option is to have the second kind of
formula where the average tariff rate of a country is
one of the coefficients in the formula. In this case
the formula could be
Interpreting for Development
15
Tl
_ (B -x-X) xT , w|lerc 3 js a coefficient,
(B xX) + T
X = Average tariff rate for developing countries.
the coefficient will be 26. However, this, as we have
seen before, cannot be accepted, as it will lead to a
very steep reduction.
The advantage of having the average tariff rate of a
country in the Swiss formula is to ensure that there is
no disproportionate tariff cut. Besides, the new tariff
rate is reflective of the tariff structure of a particular
country and is not independent of it. For instance, if
we apply the second option to the example of BTR
of textile and clothing of US and India we will get a
clearer picture. If we use this option to cut the BTR
of textile and clothing in US (T= 3.2), the new tariff
rate (Tl) will be 2.3 (X = 3.2 {Average tariff rate
of US), B = 1). This is a reduction of 73 percent.
Now, if we apply the principle of 2/3rd for India,
then India will cut its tariff rate by 48 percent, which
will result in the final tariff rate being 13 percent. In
the formula option one, with 13 as final tariff rate,
Now, if we apply the formula option two to the facts
at hand, the final tariff rate (Tl) for India will be 1 5
percent. (B = 1, X = 34.3 {Average tariff rate}). This
may be acceptable. It is important to note here that
if the value of B increases the reduction will be less.
The proposal to use average BTR of a country
as the coefficient, drew opposition from some
developing countries whose average BTR is very
low.15 These countries have argued that if they use
average BTR as the coefficient, then they will have
to undertake reductions, which may be more than
2/3rd of the reduction of developed countries. This
is a genuine difficulty. For such countries, a mark
up to the average BTR could be proposed. This
BOX 3
Mark up for Countries whose Average Bound Tariff Rate is Low
Many South East Asian countries such as Malaysia and Thailand opposed the ABI formula on the ground
that this formula will only benefit countries such as India who have a high average BTR. This opposition is
not without reasons. For instance, if we use the example ofTextile and Clothing of say, US and Malaysia,
we will find that if the average tariff rate of a country is used as one of the coefficients, then these countries
will have to undertake very steep reduction and in fact more than 2/3rd of what developed countries will
undertake and hence violate the principle of LTFR.
In the case of US, the use of ABI formula will bring down the BTR ofTextile and Clothing from 8.6 percent to
2.3 percent (reduction of 73 percent). The use of the ABI formula in the case of Malaysia (Average B3 R = 14.9
percent, BTR ofTextile and Clothing =19.5 percent) will result in BTR of Textile and Clothing coming down
from 19.5 percent to 8.4 percent. This is a reduction of 56 percent and much more than the 2/3rd reduction of
73 percent which is 48 percent. Hence, Malaysia in this case is undertaking a reduction that is much more than
warranted by the LTFR principle of 2/3rd of what developed countries undertake.
In such cases and for such countries, the average BTR should be increased by certain percentage points. This
increase will then ensure that these countries do not undertake reduction that is more than 2/3rd of what
developed countries are cutting. So, in this case if the average BTR is increased by 6 percentage points i.e.
from 14.9 percent to 21 percent, then the reduction will not be more than 2/3rd of what developed countries
are undertaking. Hence, in this particular example 6 percentage points is the mark up.
15 The average BTR of non-agricultural goods in South Asian countries is very less. For instance in Malaysia the average BTR is 14 J percent.
Similarly, the average BTR in Thailand is 24.2 percent.
16
Industrial Tariffs and South Asia
mark up could be at levels that will not make these
countries undertake tariff reductions that are more
than 2/3rd of the cuts of developed countries (See
Box 3).
The above discussion reveals that both the formula
options could be used, as they will honour
real ‘LTFR’. However from the perspective of
developing countries, the second option is better,
as it can overcome the limitation of the first
option. Moreover, having a country’s average
tariff rate as one of the coefficients in the formula
(suggested by Argentina, Brazil and India) or
having a formula that gives credit to countries
apart from incorporating the average tariff rate
(suggested by the Caribbean countries) allows
taking into account the existing tariff structure
of an individual country in the process of cutting
the tariff rates. The presence of the average tariff
rate as a coefficient will make sure that the new
tariff rate is in consonance with the existing tariff
structure of a country.
6. Eliminating Tariff Peaks and Tariff Escalation
if we apply the ABI formula to this tariff rate of
20 percent (with the average tariff in the case of
US being 3.2 percent) then, this tariff rate of 20
percent will come down to 2.7 percent.17
The other important factor that is to be used in
determining the coefficients in the Swiss formula,
according to paragraph 14, is to ensure that it
eliminates tariff escalation and tariff peak. The
elimination of tariff peak and tariff escalation
implies eliminating tariff peaks and tariff escalation
in developed countries like US and EU especially
on products of interest to developing countries. In
order to eliminate tariff peaks and tariff escalation
in developed countries it is imperative to have a
smaller coefficient in the Swiss formula, as lesser
coefficient will imply stringent reduction.
This clearly shows that when the ABI formula is
used the peak tariff of 20 percent reduces by 86.5
percent, whereas if the coefficient of 10 in the ‘Swiss
formula with two coefficients’ is used the peak tariff
of 20 percent reduces by 67 percent. Hence, the ABI
formula yields better results as far as eliminating
tariff peaks and tariff escalation is concerned.
Flere again if we apply the coefficient of 10 that
US is proposing in the ‘Swiss formula with two
coefficients’, to say a tariff rate of 20 percent in the
US, it will come down to 6.6 percent.16 However,
Hence, the above discussion clearly demonstrates
that the mandate of paragraph 14 of the HK
declaration can only be met if we follow the second
formula option for reducing tariff rates.
1'Theformula that has been usedhereis 71 ~ —, whereT 1 (final tarifFrate) = 6.6percent, T(initialtariffrate) = 20 percentandB (coefficient) = 10.
17 The formula that has been used here is 7'1 = I--------- — , where T1 (final tariff rate) = 2.7 percent, T (initial tariff rate) = 20 percent. X
(BxX) + T
(Average tariff rate) = 3.2 percent and B (Coefficient) = 1.
Interpreting for Development
17
7. Implementing the Tariff Reduction
The implementation of the tariff reduction
process is extremely important. Many have argued
that even steeper and deeper cuts can be managed
if the implementation period is long. In other
words, if countries have more time to implement
the reduction in tariffs, there will be more time
to absorb the shocks of tariff reduction. On the
other hand, if developing countries are asked to
cut their tariffs drastically or in a short span of
time, it may lead to adverse consequences such as
sudden elimination of tariff revenue or surge of
imports.
In this regard it is proposed that the implementation
period should be long and should be back loaded.
In other words, developing countries should cut less
in the initial years of the implementation period
and more in the later stages.
TABLE 8:
Hypothetical Application of the Implementation Period for a Developing Country Assuming
Hypothetical Bound Tariff Figures in Percentage
Product
X
Final
Bound
Tariff (T2)
Initial
Bound
Tariff (Tl)
14.8
100.0
Tariff After
Phase 1
(Ta)
Tariff After
Phase 3
(Tc)
Tb - X
Tariff After
Phase 4
(Td)
Tc-X
Tariff After
Phase 5
(Te)
Tl — X18
Tariff After
Phase 2
(Tb)
Ta - X
87.22
74.44
61.66
48.88
14.8
Td - Y19
Source: Author’s Calculation.
The paper proposes an implementation phase of
five stages with each stage of two equal years. The
reduction in the tariff rates by developing countries
should be spread out over all the five phases. It is
proposed that 60 percent of tariff reduction should
take place in the first four stages (equal instalments)
and 40 percent of reduction in the last phase. For
instance, assume that the initial bound tariff rate
for a product is 100 percent and this comes down
to 14.8 percent after applying the tariff reduction
formula. So the tariff reduction that has to take
place is 85.2 percentage points. Now 60 percent of
85.2 should be reduced in the first four phases i.e.
first eight years (equal instalments) and 40 percent
of 85.2 in the last phase i.e. after the completion of
the tenth year.
18 X = 60 percent of (Tl-T2)/4, where T1 is the initial bound rate and T2 is the final bound tariff.
19 Y = 40 percent of (T1-T2), where TI is the initial bound rate and T2 is the final bound tariff.
18
Industrial Tariffs and South Asia
8. Conclusion
Given the important role that industrial tariffs
play in the industrial development of a country,
it is imperative that developing countries are able
to make a judicious use of industrial tariffs. In
order to make the judicious use of industrial tariffs
it is necessary that these countries have enough
policy space. This judicious use would require
that countries have adequa+9ir tariff rates if the
situation so demands. Hence, any tariff reduction
should not eat into all the ‘water’ in the existing
tariff structure. Moreover, tariff reduction should
be a substantial process and developing countries
including the countries of South Asia should be
given enough time to reduce their tariffs. This will
not only absorb the shock effect that may be caused
by the reduction of tariffs but also provide adequate
time to the countries to develop their capabilities.
This paper has attempted to develop an
interpretation of the existing tariff reduction
modality for the adoption of the ABI or the
Caribbean formula. It has tried to show that the
only manner by which paragraph 14 of the HK
Declaration can be operationalised is to have the ABI
or the Caribbean formula to cut industrial tariffs.
A ‘Swiss formula with two coefficients’ will be in
violation of paragraph 14 of the HK Declaration.
The ABI or the Caribbean formula will also suit
the interests of South Asian countries by not asking
them to undertake stringent obligations.
However, it seems that there is a tacit acceptance by
developing countries such as India and Brazil that
a ‘Swiss formula with two coefficients’ is the way
to go ahead. Many reports from Geneva indicate
about this unfortunate development. One wonders
what is rhe reason behind this. It is important to
bear in mind that India and Brazil had proposed
the ABI formula to counter the pure Swiss formula
proposals that had been made by the EU and US.
The HK Declaration offers enough flexibility
for empowering and enabling countries to argue
for an ABI or a Caribbean type of formula. It is
up to countries like India and Brazil to continue
fighting the battle for the ABI type formula. If this
battle is not fought and won, it would indeed be
an unfortunate outcome of the multilateral trade
negotiations.
Interpreting for Development
19
Annexures
( All the tables given below are based on Author's calculations)
ANNEXURE 1
Impact of the 'Swiss Formula with Two Coefficients' on the Bound Tariff Rates of Seven Sectors
of India
The formula that has been used here is:
R xT
T\ =----- —-, where B = 5, 15, 20, 30, 35
B+T
TABLE 9
Electronics & Electrical Goods
Initial Tariff
Rate (TO)
No. Of
Tariff Lines
0
15
25
35
14
42
93
144
Final Tariff Rate (Tl)
B= 5
B= 15
B=20
B=30
B=35
0
3.75
4.17
4.38
0
7.50
9.38
10.50
0
8.57
11.11
12.73
0
10.00
13.64
16.15
0
10.50
14.58
17.50
TABLE 10
Fish and Fish Products
Initial Tariff Rate
(TO)
No. of
Tariff Lines
Final Tariff Rate (Tl)
B= 5
B= 15
B=20
B=30
B=35
0
0
0
3
0
0
0
5
9
2.50
3.75
4.00
4.29
4.38
15
6
3.75
7.50
8.57
10.00
10.50
25
32
4.17
9.38
11.11
13.64
14.58
35
397
4.38
10.50
12.73
16.15
17.50
40
1
4.44
10.91
13.33
17.14
18.67
45
4
4.50
11.25
13.85
18.00
19.69
50
1
4.55
11.54
14.29
18.75
20.59
75
2
4.69
12.50
15.79
21.43
23.86
100
23
4.76
13.04
16.67
23.08
25.93
170
1
4.86
13.78
17.89
25.50
29.02
TABLE 11
Footwear
Initial Tariff
Rate (TO)
No. of
Tariff Lines
35
Industrial Tariffs and South Asia
29
Final Tariff Rate (Tl)
B= 5
B= 15
B=20
B=30
B=35
4.38
10.50
12.73
16.15
17.50
TABLE 12
Leather Goods
Initial Tariff Rate (TO)
No. of Tariff Lines
35
22
Final Tariff Rate (Tl)
B= 15
B=20
B=30
B=35
4.38
10.50
12.73
16.15
17.50
B= 5
B= 15
B = 30
B=35
B= 5
TABLE 13
Motor Vehicles Parts and Equipments
Initial Tariff Rate (TO)
Final Tariff Rate (Tl)
No. of Tariff Lines
B=20
0
1
0
0
0
0
0
35
60
4.38
10.50
12.73
16.15
17.50
105
15
4.77
13.13
16.80
23.33
26.25
TABLE 14
Stones, Gems and Precious Metals
Initial Tariff Rate (TO)
B= 5
Final Tariff Rate 1(Tl)
B= 15
B = 20
B=30
B = 35
4.38
10.50
16.15
17.50
B= 5
Final Tariff Rate (Tl)
B= 15
B=20
B=30
B=35
No. of Tariff Lines
52
35
12.73
TABLE 15
Textiles and Clothing
Initial Tariff Rate (TO)
No. of Tariff Lines
5
1
2.50
3.75
' 4.00
4.29
4.38
15
14
3.75
7.50
8.57
10.00
10.50
20
148
4.00
8.57
10.00
12.00
12.73
25
45
4.17
9.38
11.11
13.64
14.58
30
73
4.29
10.00
12.00
15.00
16.15
35
538
4.38
10.50
12.73
16.15
17.50
Interpreting for Development 21
ANNEXURE 2
Impact of the ABI Formula on the Bound Tariff Rates of Seven Sectors of India
TABLE 16
Electronics and Electrical Goods Sector
Initial Tariff
Rate (TO)
No. of
Tariff Lines
0
14
15
25
35
Final Tariff Rate (Tl)
B = 0.5
B= 1
B = 1.5
B = 1.75
B= 2
0.00
0.00
0.00
0.00
0.00
42
8.00
10.44
11.61
12.00
12.31
93
10.17
14.46
16.82
17.65
18.32
144
11.51
17.32
20.83
22.11
23.18
TABLE 17
Fish and Fish Products
Initial Tariff
Rate (TO)
No. of
Tariff Lines
n
Final Tariff Rate (Tl)
B = 0.5
B =1
B = 1.5
B = 1.75
B=2
0
3
0.00
0
0
0
0.00
5
9
3.87
4.36
4.56
4.62
4.66
15
6
8.00
10.44
11.61
12.00
12.31
25
32
10.17
14.46
16.82
17.65
18.32
35
397
11.51
17.32
20.83
22.11
23.18
40
1
12.00
18.47
22.50
24.00
25.27
45
4
12.42
19.46
24.00
25.72
27.17
50
1
12.77
20.34
25.36
27.28
28.92
75
2
13.96
23.54
30.52
33.34
35.83
100
23
14.64
25.54
33.97
37.51
40.69
170
1
15.58
28.54
39.50
44.36
48.88
TABLE 18
Footwear
Initial Tariff
Rate (TO)
35
Final Tariff Rate (Tl)
No. of
Tariff Lines
29
Industrial Tariffs and South Asia
B = 0.5
B= 1
B =1.5
B = 1.75
B=2
11.5
17.3
20.8
22.1
23.1
TABLE 19
Leather Goods
Initial Tariff
Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
MEM
TABLE 20
Motor Vehicles Parts and Components
Initial Tariff
Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
B = 0.5
B=1
B = 1.5
B = 1.75
3= 2
0.00
0
1
0.00
0.00
0.00
0.00
35
60
11.5
17.3
20.8
22.1
23.1
105
15
14.7
25.8
34.5
38.1
41.4
TABLE 21
Stones, Gems and Precious Metals
Initial Tariff
Rate (TO)
No. of
Tariff Lines
35
52
Final Tariff Rate (Tl)
B = 0.5
B= 1
B = 1.5
B = 1.75
B= 2
11.5
17.3
20.8
22.1
23.1
TABLE 22
Textiles and Clothing
Initial Tariff
Rate (TO)
5
Final Tariff Rate (Tl)
No. of
Tariff Lines
J
B = 0.5
B=1
B = 1.5
B = 1.75
B= 2
1
3.87
4.36
4.56
4.62
4.66
15
14
8.00
10.44
11.61
12.00
12.31
20
148
9.23
12.63
14.40
15.00
15.49
25
45
10.17
14.46
16.82
17.65
18.32
30
73
10.91
16.00
18.95
20.00
20.87
35
538
11.51
17.32
20.83
22.11
23.18
Interpreting for Development ~23
ANNEXURE 3
Impact of the Caribbean Formula on the Bound Tariff Rates of Seven Sectors of India
71 = y—————~, where Tl is the final tariff rate, T is the initial tariff rate, B = 1 (a coefficient),
{(d + C) xa/ + 1
X = 34.3 percent (Average tariff rate of a country) and C = 1, 2, 3, 4 (credit to be accorded to a developing
country).
TABLE 23
Electronics & Electrical Goods
Initial Tariff Rate (TO)
No. of
Tariff Lines
Final Tariff Rate (Tl)
C= 1
C= 2
C=3
C=4
0
0
0
0
14
0
15
42
12.31
13.09
13.52
13.79
25
93
18.32
20.11
21.15
21.82
35
144
23.18
26.12
27.89
29.07
TABLE 24
Fish and Fish Products
Initial Tariff Rate (TO)
No. of
Tariff Lines
C=1
C= 2
C=3
C=4
0
3
0
0
0
0
4.77
4.82
4.86
Final Tariff Rate (Tl)
5
9
4.66
15
6
12.31
13.09
13.52
13.79
25
32
18.32
20.11
21.15
21.82
35
397
23.18
26.12
27.89
29.07
40
1
25.27
28.80
30.97
32.43
45
4
27.17
31.31
33.89
35.65
50
1
28.92
33.65
36.65
38.71
35.83
43.38
48.49
52.18
75
2
100
23
40.69
50.71
57.84
63.17
170
1
48.88
64.10
75.92
85.37
Initial Tariff Rate (TO)
No. of
Tariff Lines
TABLE 25
Footwear
24
Industrial Tariffs and South Asia
Final Tariff Rate (Tl)
TABLE 26
Leather Goods
Initial Tariff Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
■ms Kia
TABLE 27
Motor Vehicles Parts and Equipments
Initial Tariff Rate (TO)
Final Tariff Rate (Tl)
No. of
Tariff Lines
C=1
C=2
C=3
C=4
1
0
0
0
0
35
60
23.18
26.12
27.89
29.07
105
15
41.49
51.97
59.48
65.13
0
TABLE 28
Stones, Gems and Precious Stones
Initial Tariff Rate (TO)
No. of
Tariff Lines
35
52
Final Tariff Rate (Tl)
C= 1
C= 2
C=3
C=4
23.18
26.12
27.89
29.07
TABLE 29
Textiles and Clothing
Initial Tariff Rate (TO)
No. of
Tariff Lines
5
1
4.66
4.77
4.82
4.86
15
14
12.31
13.09
13.52
13.79
20
148
15.49
16.75
17.46
17.91
25
45
18.32
20.11
21.15
21.82
30
73
20.87
23.23
24.62
25.53
35
538
23.18
26.12
27.89
29.07
Final Tariff Rate (Tl)
C=1
C= 2
C=3
C=4
Interpreting for Development
25
Centre for Trade and Development (Centad) is an independent,
not-for-profit think tank that carries out policy research and advocacy
on issues around trade and development, with a principal focus on
South Asia.
Centre for Trade & Development
# 406, Bhikaiji Cama Bhavan
Bhikaiji Cama Place
New Delhi -110066
Tel: + 91 -11 -41459226
Fax: + 91 -11 -41459227
Email: centad@centad.org
Web: www.centad.org
Position: 4657 (1 views)