Challenges in estimating effects of trade reform

1
Challenges in estimating effects of trade reform

• Kym Anderson
• Development Research Group, World Bank
• Washington DC

2
Estimating ex ante effects of trade reform

• Global trade modeling began to make an impact
  from the early 1980s
• In agriculture, initially PE models were popular
  for getting price, trade and welfare effects
• Starting with single-commodity, then
  multi-commodity
• Okay for rich countries where agriculture is a
  small part of the economy and manufacturing
  tariffs are low, but otherwise welfare
  calculations are unreliable
• Unable to show inter-linkages between agric and
  other sectors, as needed in UR multilateral trade
  negotiations
• CGE national and then global models (esp. GTAP)
  have since become the main workhorse for MTNs

3
CGE models have come a long way

• National models e.g. Evans (1972) for Australia,
  leading to comparative static ORANI (1982) and
  then dynamic MONASH (1998)
• Global models e.g. Whalley (1985) and Deardorff
  and Stern (1986, 1990), then SALTER (1991) which
  led to GTAP
• GTAP global database now Version 6, with 57
  sectors (20 agr.) and 87 countries/regions
• each with Armington elasticities (to capture
  bilateral and intra-industry trade better)

4
GTAP database, Version 6

• Based on 2001 market and policy data, it is
  superior to earlier versions because it has
• Bound as well as applied tariffs
• AVEs of specific and compound tariffs
• TRQs accounted for (if not very well)
• Non-reciprocal (EBA, AGOA, CBI, ACP) as well as
  reciprocal (FTA) tariff preferences
• Helps address concerns of Panagariya (2005)

5
Where to from here?

• Improve the specification of existing and
  alternative policy instruments
• NTBs such as SPS and TBT measures, including new
  ones such as GMOs
• Creating a two-tier trading world? (Baldwin 2001)
• Safeguards and contingent protection
• e.g. continued use of agric variable levies by
  EU, Chile while bound tariffs remain well above
  applied rates
• Export taxes/restraints (including contingent, as
  with EU wheat in 1995, China textiles from 2005?)
  
• protection re-instrumentation
• Services trade and investment distortions

6
Where to from here?

• Improve the specification of existing and
  alternative policy instruments (cont.)
• Better representation of agricultural TRQs and
  other non-reciprocal preferences
• Including who gets rent
• Quota underfill due to ROOs administration
• Blue-box supply restraint measures
• Green-box measures not fully decoupled

7
Where to from here? (continued)

• Specify other domestic divergences
• Product, factor and income taxes/subsidies
• Labour union interventions
• Environmental and social concerns (whose?) that
  are not being addressed?
• But beware of excuses for intervention (e.g.
  agricultures so-called multifuntionality)
• What is the appropriate counterfactual?
• Its not necessarily the status quo

8
Where to from here? (continued)

• Imperfect competition and scale economies
• EU modellers have made most progress here, as it
  has been a major contributor to gains from moving
  to a single EU market
• Will be increasingly important as modelling
  expands of the service sectors trade and
  investment distortions
• e.g. distribution services incl. retailing
  (Walmart)

9
Where to from here? (continued)

• Dynamics of trade liberalization potentially can
• Integrate growth and trade modelling
• Allow asset market to be included and hence
  intergenerational effects to be estimated
• Allow closer scrutiny of the options for
  developing countries in accommodating to a
  reduction in tariff revenue (e.g. via foreign
  debt financing)
• Show how much greater are the gains from trade,
  and how small are the adjustments to reform
  compared with normal adjustments to growth
• Politicians want ests. of short-run adjustment
  effects

10
Where to from here? (continued)

• Effects on factor markets, esp. labour
• Not often done because of simplistic CGE
  modelling of factor markets?
• Need to improve so as to get more realistic
  effects of trade reform, and also to feed into
• Effects on household and regional income
  disparities and poverty alleviation, by
  integrating survey data into CGE models
• major concordance challenges are being overcome

11
Where to from here? (continued)

• Domestic policy responses to trade reform
• Positive analysis what might governments do to
  domestic policies if trade reform is introduced?
• Normative analysis what should governments do in
  terms of domestic policies to more-efficiently
  satisfy societys objectives (poverty
  alleviation, income distribution, etc.) if/when
  trade reform threatens them?

12
Has better measurement of effects led to more
trade reform?

• How much credit is due to economists?
• Australias major reforms transformed its
  economy from almost slowest- to fastest-growing
  in OECD?
• Analysis of East Asias export-led growth?
• Agricultures inclusion in the UR and DDA, and
  prioritization of measures?
• See next two tables

13
(No Transcript)
14
(No Transcript)
15
What are the highest priorities for trade policy
analysts now?

• Better estimates of distortions to agricultural
  and services markets
• Better understanding of the political economy
  process
• Build relations with policymakers their
  advisors
• Address the empirical questions in the minds of
  policy makers/advisors/NGOs
• evaluate alternative measures, including buyouts?
  
• Become more-effective advocates for liberal
  trade? e.g. in disseminating the following Doha
  simulation results

16
Key elements of the Doha Agenda

• Agric market access is by far the biggest
  potential contributor to global and
  developing-country welfare gains
• We assume no services reform, but we model the
  following
• Phase-out agricultural export subsidies
• Reduction of agricultural domestic support
• Cuts in agric and non-agric bound tariffs under
  various alternative market access packages

17
Agricultural market access

• Tiered formula for cutting bound tariffs (with
  Special and Differential Treatment for developing
  countries, or SDT)
• allow a few sensitive products to be exempt
• We compare it with a proportional cut of same
  average size for high-income and developing
  countries
• And we look at effects of imposing a tariff cap
  of 200

18
Agricultural domestic support

• Cut in bound AMS need not reduce applied support,
  because of binding overhang (with 1986-88 ref.
  prices)
• and overhang can be increased by abolishing admin
  prices used to calculate MPS
• Removes MPS from current AMS, but leaves it in
  binding
• We apply a tiered reduction to the bound AMS 75
  if AMSgt20, otherwise 60
• Leads to only 5 countries reducing support
• Thailand 30, US 28, Norway 18, EU 16,
  Australia 10

19
Other bits of the Doha package

• 50 cut in bound tariffs on non-agricultural
  goods for high-income countries and 33 for
  developing countries (except zero for LDCs)
• We also examine the effects of developing
  countries becoming full participants in Doha
  (foregoing Special and Differential Treatment)
• recalling from earlier GATT Rounds that
  developing countries only got what they gave, in
  terms of increased market access
• Finger (1974, 1976, 2001)

20
Results from Doha agric reform

• Tiered formula cut in tariffs with SDT gives the
  world 72 billion per year
• but only 7 billion go to developing countries
• and if rich countries exempt just 2 as sensitive
  products, global gain shrinks to 16 billion and
  developing countries lose
• although a 200 tariff cap reduces much of that
  shrinkage
• If tiered formula is replaced by a proportional
  one to yield same average tariff cuts, global
  gain is nearly as high (64 billion)
• ? Worth bothering to negotiate a complex tiered
  formula? Or use a Swiss formula instead?

21
Adding non-agric market access

• boosts global gain from 72 to 95 billion
• which gets the world 1/3rd of the way to the
  potential gains from complete free trade in
  merchandise
• If developing countries were to forego SDT in
  market access, global gain would rise from 95 to
  119 billion per year
• and developing countries gain would rise by 50,
  to 23 billion per year
• and US gain would nearly double, to 7 billion

22
Implications for developing countries

• Most of their gains go to large agricultural
  exporters, while low-income Sub-Saharan African
  countries gain little
• To gain more, poor countries have to liberalize
  more
• which they would be better off doing under Doha,
  so as to get reciprocity (greater market access
  abroad) and/or more aid, than unilaterally?

23
DC welfare gains with SDT (percent change
from baseline income in 2015)
24
DC welfare gains without SDT(Percent
change from baseline income in 2015)
25
DC gains from full global libn(Percent
change from baseline income in 2015)
26
Conclusions and policy implications

• Potential gains from further trade reform are
  large
• Even after UR and recent accessions to WTO and EU
• ?Must find the political will for Doha success
• DCs would gain disproportionately from reform
• Notwithstanding non-reciprocal tariff preferences
• But as much would come from South-South as
  South-North trade growth, hence the importance of
  developing countries own liberalization
• After outlawing export subsidies, agric tariff
  cuts are the highest priority from a welfare
  viewpoint and if Doha is to be pro-development/pro
  -poor

27
Conclusions and implications (cont)

• Cuts in agric tariffs and domestic support
  bindings need to be large to get beyond binding
  overhang
• Even large cuts in agric tariffs do little if
  sensitive and special products are exempt
• Unless a tariff cap of, say, 100 is enforced
• DCs would have to make few cuts because of their
  huge binding overhang
• So they can afford to tone down their demands for
  SDT (and special products) and trade it for
  greater access to rich-country markets ( fewer
  sensitive product exemptions by rich countries)