Transcript Aucun titre de diapositive - Organization of American States

Political economy
SESSION IV
THE POLITICAL ECONOMY OF
REGIONALISM AND IMPLEMENTATION
ISSUES
JAIME DE MELO
This version, March 2004
1
Political economy
OUTLINE: SESSION IV
 Elements of the political economy of protection
 Are government commitments credible?
 Is regionalism likely to increase or diminish
protection towards the outside world?
 Application: The CET in the MERCOSUR
 Regionalism as politics
 Implementation issues : Rules of Origin
 Application: Rules of Origin in NAFTA
2
Political economy
 Elements of the political economy of protection
General framework
Individual or group
preferences:
factor income, cost
of living, etc..
Box A
Pressure groups:
Lobbying to get policy
(more or less
protection); type of
protection (e.g. QRs)
“Demand”
for/against
protection and
type of protection
Box B
Choice of
commercial policy
Politicians ’
preferences:
Institutional set up:
Level, type (QRs?)
of protection
(executive vs. legislative;
international obligations -WTO, RTA, WB)
Box C
Box D
“Supply”
for/against
protection and
type of protection
3
Political economy
 Elements of the political economy of protection
BASIC QUESTIONS ON THE FRAMEWORK
Within the general framework:
1. Who are the lobbies and how do they work (box B)?
2. What is the institutional setting (box D)?
 To make framework more operational in N-S RIA setting:
3. Are commitments more credible in an RIA?
4. Elements of demand and supply of protection
4
Political economy
 Elements of the political economy of protection
WHO ARE THE LOBBIES?
 Industrialized countries:
• Most powerful lobbies represent producers (eg.
guns, steel, automobile, farmers,…)
• Few powerful lobby groups representing
consumer interests
 Developing economies:
• Producers and importers often represent
powerful lobbies
• Few lobbies representing rural areas, especially
in low-income countries
5
Political economy
 Elements of the political economy of protection
HOW DO LOBBIES WORK?
 Public good characteristic of lobbying activity:
Public good: non rival (several consumers at a time)
and non excludable (free riders cannot be excluded).
Ex: A tariff on shirt imports obtained by producer
i benefits all national shirt producers = protection
(as free trade and many other policies) has a
public good nature
Lobbies face free-riding = incur monitoring costs
Monitoring costs are usually higher the larger the
number of participants in the group = most
efficient lobbies tend to be the less numerous
ones = “tyranny of the many by the few”.
6
Political economy
 Elements of the political economy of protection
THE INSTITUTIONAL SET-UP
 Internal institutions:
How are rules and laws decided and implemented?
“Checks and balances” limits on discretionary action?
= voting process, lobbies contributions,
separation of power, law enforcement, etc...
 External institutions:
= membership in WTO or in RTAs?
May benefit countries with weak institutions as an
“external agency of restraint”, putting a lid on the amount
of discretionary action in trade (and other) policies
7
Political economy
 Are government commitments credible?
Often policies that appear optimal ex-ante no longer are
so ex-post = there is a problem of “time-consistency”
Time-consistency problem: Particularly relevant when
decisions, once taken are costly to undo and the
economy has rents to be fought over= LDCs
Ex.: excessive taxation of crops with high sunk costs:
Growers of tree crops (eg coffee) consider the use of
fertilizers as sunk costs.
 government has incentive to cheat: announce that taxes
will be low, to induce producers to use fertilizers, that
increase their profits and thus the tax base, … and then
keep taxes high.
../..
8
Political economy
 Are government commitments credible (II)?
But: growers know from previous experience that
government will cheat  no fertilizers applied 
tax base shrinks and everybody is worse off.
Data show that crops with high ratio of sunk costs
(coffee, vanilla, cocoa) are taxed at a higher rate
than other crops (cotton, tobacco, groundnuts).
So are trade policies credible or do governments
have incentives (and leeway) to renege its promises?
9
Political economy
 Are government commitments credible? (III)
Will N-S RIAs bring credibility to Southern countries?
Here role for RIAs (eg. the EA between transition
countries and the EU, NAFTA, REPAs) that can serve as
external agencies of restraint to limit discretionary power
of government, or backsliding in policies
 Probably important in some cases where that
credibility role would not be played by the WTO (notably
because the countries would not avail themselves with
binding tariffs at actual levels)
….but is that sufficient if the Northern partner does not
have that much interest in the Southern partner?
…and is it a diversion of scarce human capital for more
negotiations?
10
Political economy
 Regionalism and protection towards the outside world
Do RIAs change the balance of incentives and forces for
external liberalism since the gains from an RIA depend on a
genuine liberal external policy stance?
 Compare first pressures under FTA and under CU
 Keep in mind: No simple answer, as there is no empirical
regularity: The CAP is protectionist (but would individual
countries have been on average more or less protectionist?)
Under NAFTA, Canada reduced tariffs on 1500 items
imported from non-partners, while Mexico raised tariffs on
503
 EC: Average CET has fallen from 13% to 3% for
manufactures
CET on Andean Pact at 12.8% percent just above the
average tariff of members.
11
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in FTAs
A race to the bottom for FTAs
• Optimal tariff on non-members should be close to zero if
tariffs on members are constrained to zero
• If there is trade deflection, high tariff countries lose
revenue, unless they reduce tariff just below that of the
partner (à la Bertrand competition)
• If duties on inputs cannot be rebated (often part of the
agreements), then exporters of final goods are
uncompetitive.
12
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in FTAs (end)
…but counteracting forces to the race to the bottom for
FTAs
• RoO essentially export protection and there is plenty of
lobbying for obtaining RoO that allow exporting protection
to the partner.
• Fierce competition with the FTA might lead firms to seek
protection against outsiders
• And resources for lobbying previously devoted to
lobbying for protection against partners may now be freed
to seek lobbying against non-partners.
13
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in Customs Unions (I)
Most CUs have a CET but not a truly common external
trade policy (e.g. the EU-Turkey CU allows for AD between
members.
…took 30 years for the EU to get to common external trade
policy as for a long time members could keep their own
NTBs towards outside world (so RoO were indeed still
needed).
• If CET is more uniform than previous regime, less
lobbying and extra gain at same average level of protection
• By coordinating their policies CU members gain power
and hence may be tempted to use it.
• Large countries usually dominate (e.g. Brazil), and since
large countries usually have higher protection, the CET
may be higher as a result.
14
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in Customs Unions (II)
CUs require the creation of supranational institutions. Are
these going to increase or decrease protectionist pressures?
Suppose that tariffs are strategic substitutes in the strategic
interactions among blocs (if one blocs increases its tariff,
then it is optimal for the other bloc to lower its tariffs ---as
quantities in a Cournot game). → Pays to delegate and let the
more aggressive member negotiate (but the opposite if tariffs
are strategic complements).
► Not clear whether delegation leads to higher or lower tariffs
in a CU
► Moreover, since negotiation is a continuous process
(repeated game), probably a more aggressive leader will be
able to extract better terms because of more credible threats
●
15
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in Customs Unions (III)
Typically (like the EU), a disproportionate influence is given to
countries in areas in which they have a vital interest (i.e. to
producer interests). So if comparative advantage is at work,
over time the powerful sectors should be those that are
efficient, thereby reducing protectionist pressures.
…But biases towards protectionism still there
► More layers of bureaucracy shift the influence away from
voters towards producers.
► with costs of protection more thinly spread and benefits
concentrated, bias towards producers and protection.
16
Political economy
 Regionalism and protection towards the outside world :
Protectionist pressures in Customs Unions (end)
---two institutional failures
► restaurant bill problem (bigger bill when split evenly).
Suppose benefit of a policy is proportional to its share in CU
output, while costs are proportional to its share in GDP (a
tariff which benefits producers at the expense of consumers
has this characteristic). Countries around the negotiating
table will each press for protection for goods where their
share is higher than their GDP share. →All measures are
added even though individually each country would have
preferred no change.
►Universalism: benefits go to one country (say there are
three countries each producing a type of steel: spoils go to
the country that gets it, while costs are borne by all. The worst
outcome is if protection passes but not for the type of steel
you produce…so you back the protectionist measure.
17
Political economy
 Regionalism and protection towards the outside world :
Lobbying activities in RIAs
Examples: Citrus growers in US oppposed FTAA and got 15
years to adjust to NAFTA (among most heavily protected
sectors…..)
UEMOA CET in 1998: 5% on interm., 10% on capital goods,
20% on consumer goods.
…debate led to cement being declared a consumer good!!!
Problem for gv’ts : TC hurts the most powerful lobbyists
usually in import-competing activities.
Brazil wanted free trade in IT for MERCOSUR, but was a
holdout on the nearly global IT Agreement of 1996.
 see the following model (due to P. Krishna) where trade
policy in a three-country world is entirely determined by
profits
18
Political economy
 Regionalism and protection towards the outside world :
Lobbies bias RIAs towards Trade Diversion (TD)
Take two countries where oligopolists (making a positive
profit) play Cournot. Then producers in A and B might well
support an FTA because the gain they would make in each
other’s country (notably at the expense of producers in C)
would exceed the losses they would make via reduction in
market share to firms in partner country
…firms likely to support a TD FTA!!!
 What about incentives for firms to participate in a MTL
after the FTA is formed? Most unlikely as the MTL will be
essentially giving market share, and hence profits to firms
in C
 FTAs can well be a stumbling bloc in the road to
reaching multilateral free trade!!!
19
Political economy
 Regionalism and protection towards the outside world :
RIAs: A new environment for lobbying?
Preference dilution effect. One view (see MP chapter 6 is
that an RIA has a preference dilution effect because there is
more opposition to overcome, more of a free-rider problem,
more representatives to influence)
Assumption here is that fragmented lobbies face a unified
CU government. But what about the opposite?
See the discussion in SW pp. 92-3 and the discussion about
the growth of lobbying 300 in 1970 to 3000 in 1990) activity
in Brussels.
 In conclusion much depends on what type of institutions
are adopted.
20
Political economy
 Application: The CET in the MERCOSUR S-W box 3.3
With its highly non-uniform Common External Tariff—with
numerous exceptions—Mercosur is an ideal case study of the
role of pressure groups in RIAs
It took several years to negotiate Mercosur’s CET, culminating
in the December 1994 Ouro Preto Protocol. Each member was
allowed an exceptions list and the initial CET applied only to
about 75% of the universe of 9,119 tariff lines (table). Left out
were capital goods, computer and telecommunication
equipment, and automobiles and sugar.
Convergence to the CET should be achieved for most goods by
2006; there is no agreement on convergence dates for sugar
and automobiles. The Ouro Preto Protocol also established a
list of deviations from intra-Mercosur free trade. These were
planned to disappear by 2000.
21
Political economy
 Application: The CET in the MERCOSUR
Olarreaga and Soloaga tried to explain variations and
deviations from the CET across 27 industries, using measures
of political and interest group activity such as wages, industrial
concentration indices, labor/capital ratios, import penetration,
intra-industry. Main conclusions:
 Sectors where significant trade creation is likely tend to be
exempted from internal free trade. Internal free trade is also
resisted more successfully by sectors with high employment
shares. This reflects government desires to avoid large-scale
labor adjustments, and the voting strength and trade union
presence of large sectors. High third-country import
penetration, which puts competitive pressure on domestic
firms, also leads to high protection.
22
Political economy
 Application: The CET in the MERCOSUR (end)
 Political considerations explain 58% of the variation in the
CET across industries. The negotiated CET in any sector
correlates with the share of capital remuneration in value
added in that sector and the share of sector-specific capital in
total inputs. If labor is fairly mobile and new firm entry difficult,
it is existing capital that receives most of the benefits of
protection. They also found a positive correlation with industry
concentration: more concentrated industries find it easier to
organize lobbying efforts, making them more effective
lobbyists.
·  The CET in a sector mainly reflects the preferences of the
member country that has the greatest production in that sector.
Thus, Brazil’s preferences are the main determinants of the
structure of the CET, as it represents at least 70% of Mercosur
production in each of the 27 sectors considered.
23
Political economy
Regionalism as politics
(See WB chp. 1 SW chp. 7)
 An RIA can reduce frictions along several dimensions
Trade can promote peace with secure trade relations
reducing likelihood of war and RIAs promote trade….
(see EC case; MERCOSUR emerging from military
regimes)
 Will Euro-Med help diminish spread of civil
disturbance?
 Rich countries want to stem migratory pressures from
the South. But NAFTA accompanied by decrease in
wages of unskilled Mexicans…
24
Political economy
Regionalism as politics (c’td)
 Regional integration can reduce frictions along several
dimensions
Trade can promote peace with secure trade relations
reducing likelihood of war and RIAs promote trade….
(see EC case; MERCOSUR emerging from military
regimes)
 Will Euro-Med help diminish spread of civil
disturbance?
 Rich countries want to stem migratory pressures from
the South. But NAFTA accompanied by decrease in
wages of unskilled Mexicans…
25
Political economy
 Implementation issues : Rules of Origin (RoO)
RoO are necessary to prevent trade deflection (i.e.
importing from the low tariff partner in an FTA)
 Do you need RoO in a CU? In principle no, but then
ALL of trade policy should be common (e.g. in the EU
it took 30 years to get there since Article 115 of Rome
Treaty allowed contries to keep their NTBs at national
level notably in connection with the MFA  RoO were
necessary)
 But RoO do not prevent indirect trade deflection (i.e.
exporting protection as the low tariff partner imports
all its consumption needs from ROW while it exports
all its inefficient production to the high tariff partner)
26
Political economy
 RoO and governance
RoO occupy 81 pages in EU’s agreement with Poland
and 200 pages in NAFTA
Real problem, especially for low income developing
economies as it :
 Uses resources (customs officials can spend 25% of
their time or more on RoO)
 gives opportunity for discretion and bribes
 Also a political economy issue as in typical N-S
FTA, because it is the Northern country partner that
determines the rules (see NAFTA case study below)
27
Political economy
 An example of protectionist RoO
NAFTA’s RoO for catsup (S-W: Box 3.1) .
To Chile’s dismay, the tomato catsup rules changed when
CUSFTA evolved into NAFTA. Under CUSFTA, catsup
processed out of imported tomato paste qualified for dutyfree treatment in internal trade. Under NAFTA rules the
tomato paste itself must be produced within a NAFTA
member in order for the catsup to qualify for free entry. In
1992, Chile was the leading foreign supplier of tomato paste
to the United States, and the US-produced catsup enjoyed
free entry under CUSFTA. Mexico and Chile together
accounted for over 80% of US tomato paste imports in
roughly equal quantities. Under NAFTA, catsup made out of
Chilean paste could no longer circulate duty-free, Chile’s
share dropped to 5%, with Mexico taking the other 75% 28
Political economy
Viva la pasta!!!
Figure : Regional Organizations in Southern and Eastern Africa
CBI CBI CBI
BIBI
Burundi
Rwanda
COMESA
ccOMESA
COMESA
SADC
Djibouti
Egypt
Eritrea
Ethiopia
Sudan
Malawi
SADC
Angola
Dem. Rep.
Congo
Zambia
Zimbabwe
Tanzania
EAC
Kenya
Uganda
IOC
SACU
Namibia
Lesotho
EAC
Swaziland
IOC
Mauritius
Seychelles
Comoros
Madagascar
South Africa
Botswana
Mozambique
29
Political economy
Two criteria for determining origin:
• “wholly-obtained or produced” for
agricultural products
• “substantial transformation” criterion for
manufactures
Components of RoO at the product level:
1. Change of tariff classification (CTC)
2. exception attached to a particular CTC (called
an ECTC)
3. Value Content (VC)
4. Technical requirement (TECH)
See variety of criteria used by PANEURO and NAFTA in next slide!
30
Requirement
NC
NC+ECTC
NC+TECH
NC+ECTC+TECH
NC+VC
NC+ECTC+VC
NC+VC+TECH
NC+WHOLLY OBTAINED CHAPTER
NC+WHOLLY OBTAINED HEADING
SUBTOTAL
CI
CI+ECTC
CI+TECH
CI+ECTC+TECH
CI+VC
CI+ECTC+VC
CI+VC+TECH
SUBTOTAL
CS
CS+ECTC
CS+TECH
CS+ECTC+TECH
CS+VC
CS+ECTC+VC
CS+VC+TECH
CS+ECTC+VC+TECH
SUBTOTAL
CH
CH+ECTC
CH+TECH
CH+ECTC+TECH
CH+VC
CH+ECTC+VC
CH+VC+TECH
CH+ECTC+VC+TECH
SUBTOTAL
CC
CC+ECTC
CC+TECH
CC+ECTC+TECH
CC+VC
CC+ECTC+VC
CC+VC+TECH
CC+ECTC+VC+TECH
SUBTOTAL
TOTAL
EUROPE
PANEURO
0.39
2.39
1.39
0.00
11.46
1.57
0.08
7.62
0.70
25.60
AMERICAS
NAFTA
ROO INDEXa
0.54
1
1-2
2
2
4-5
5
7
7
0.54
0.02
0.02
0.00
0.20
0.00
1.90
0.00
0.27
0.00
0.00
0.00
2.37
32.99
4.60
0.00
6.66
13.01
0.37
0.00
0.02
57.65
2.16
1.02
0.04
11.02
0.00
0.00
0.00
0.00
14.24
100
0.04
1.29
2.52
0.04
0.40
0.10
4.35
17.09
19.18
0.02
0.14
3.54
0.58
0.10
40.65
30.95
17.71
0.02
5.76
1
2
2
2
2
2
3
3
3
3
Political economy
ri index takes a value
1<ri<7
See definition below
in NAFTA case
study
ri=1 (not restrictive)
ri=7 (restrictive)
4
4
4
4
5
5
5
5
6
6
6
6
7
7
7
7
54.44
100
31
Political economy
 Economics of RoO (I)
K
Fig.1 : Physical content RoO and costs
L/K=0
R
RoO: 0 1
L/K=1
R() =
Restricted
costs
1/0
O
K= imported
inputs from
ROW
A
B
L
32
 Economics of RoO (I)
Political economy
This slide explains following figure 2 (from Krishna paper)
 A & B form an FTA, both importing final good from ROW
P* =world price (taken as given). B produces the good at a constant
cost CB and has no tariff: tB= 0
tA = A’s tariff on the import and for simplicity does not produce the
good
PA and PB= (=P*=CB) denote domestic prices en A and B;
 With RoO, B firm makes profit (R)=PA - RB()
Without RoO, B firm makes profit (t)=PA - (1+tA) CB
(assuming that tariffs are levied on the price which equals domestic cost)
Firm in B will take FTA option if RB()< CB (1+tA)
33
Political economy
 Economics of RoO (II)
This slide explains following figure 2 (from Krishna paper)
For   0  R()= CB (RoO is not binding and has no
additional effect on costs)
 welfare in A increases because area DEFC (CS gain)
exceeds area EFGD (loss of tariff revenue)
For  > 0  R()> CB
And for 0 <  < 2 it is still worth it for B to use FTA.
However, as   , marginal gains and marginal costs
  =3 makes A indifferent, since FGI=KIGD. But beyond
that value A loses.
Also note that if B cannot satisfy A, it will get rents which
can explain why FTA generates FDI!!
34
Political economy
Figure 2: Demand, Price and Cost in and FTA
Price
A
F
E
CB(1+tA)
K
CB
I
RB()
J
H
D
C
G
0
3
2
B
Quantity
35
Political economy
Application: Nafta RoO Map and cost estimates
Preferences & utilization rates under NAFTA.
3 main types of RoO in NAFTA (Estevadeordal, 2000).
• A change in tariff classification at the chapter (CC),
heading (CH), sub-heading (CS) or item (CI) level;
• A domestic content rule or a regional value content
(RVC);
• A technical requirement on product and/or on process
(TECH).
36
Political economy
Preferences and Utilization Rates (c’td)
• Data on US imports from Mexico for 2001 at the
HS-6 level of aggregation, 20 sectors, 99
chapters, 3555 tariff lines ;
• ui = utilization rate = ratio of US imports from
Mexico under US-NAFTA preferential tariffs to
total US imports from Mexico;
• Tariff preference margin;
ti   i
i 
;
1  i
t  t
i
us
i ,mfn
; i  t
us
i ,mex

37
Political economy
Preferences and Utilization Rates (c’td)
Estevadeordal’s (2000) observation rule:
y = 1 if y ≤ CI
y = 2 if CI < y* ≤ CS
y = 3 if CS < y* ≤ CS and RVC
y = 4 if CS and RVC < y* ≤ CH
y = 5 if CH < y* ≤ CH and RVC
y = 6 if CH and RVC < y* ≤ CC
y = 7 if CC < y* ≤ CC and TECH
y* :latent level of restrictiveness of RoO (as
opposed to observed level of restrictiveness);
38
Political economy
Preferences and Utilization Rates (c’td)
Table 1 : RoO map, Preferences and Utilization Rates.
Obs
X to
US
%
ui
%
τi
%
CC
%
CH
%
CSI
%
E
%
TECH
%
RVC
%
ri
%
Total
3555
100
100
58,0
4,1
50,0
45,1
5,0
47,0
8,6
4,9
5.1
Raw
330
9,3
9,4
34,2
1,8
95,2
4,5
0,3
10,3
0,9
0,0
5.9
Interm 1048 29,5
4,1
74,2
4,8
58,4
39,4
2,2
68,4
0,2
8,7
5.2
Final
2177
61,2
86,6
53,9
4,2
39,1
53,9
7,0
42,2
13,9
3,8
4.9
Textile
618
17,4
7,4
79,9
10,4
80,3
19,7
0,0
97,9
41,6
0,0
6.0
All data on RoO refers to percentage of tariff lines subject to the
corresponding RoO.
39
Political economy
Preferences and Utilization Rates (c’td)
Figure 1: cumulative distribution on total sample.
Utilization Rates
Tariff Preference Rates
40
Political economy
Non-parametric Estimates
Method of revealed preference as in Anson et
al.(2004) but for 2001 data.
• For products with ui=100%, τi > ci;
• For products with ui=0%, τi < ci;
• For 0%<ui<100% . Assumption: firms are indifferent
to export to US under NAFTA or MFN regimes.
c=τ=6.16% for 2001 [vs. 6.11% for 2000]
with τ computed on 0%<ui<100%.
41
Political economy
Non-parametric Estimates (c’td)
Compliance costs: administrative component,  , and a
i
distortionary component  i :
ci   i   i
Assumption: administrative costs negligible for firms on
their participation constraint (0%<ui<100%) when ri≤2
(values corresponding to CH).
Calculating the average preference margins for
utilization rates close to 100% (say ui=95%) when ri≤2
gives an upper bound of the distortionary component,  i
or 2000 [2001].
  4.30%   4.44%
42
Non-parametric Estimates (End)
Political economy
Summary: averages for 2000 [2001]:
• Total compliance costs: c =6.11% [ 6.16%];
• Distortionary costs:  =4.30% [ 4.44%];
• Administrative costs:
  6.11%  4.30%  1.81%   6.16%  4.44%  1.72%
Administrative cost estimate for 2001
less than for 2000 (both in absolute terms
and in relative terms), as it falls from 45% to
42% of the total compliance costs. Learning!
43