DO DIFFERENCES IN LEVELS OF
PESTICIDES' MRLS AFFECT TRADE: THE CASE OF
POME FRUITS
F. DEMARIA & S. DROGUÉ
(INRA-AGROPARISTECH, FRANCE
UNIVERSITY OF CALABRIA, ITALY)
PARIS JUNE 23-25 2010
FERENTILLO JULY 8-9 2010
Background
Objectives of the study
• Assess the impact of differences in the
MLRs of pesticides on pome fruits
(apples and pears) and related
processed products on trade
competitiveness
Why Pome fruits ?
• Most consumed fruit in the US & in Europe
(with oranges)
• Unlike other fruits, easily shipped
• Temperated area product -> main
producers are developed and emerging
countries then they are less price
differences
Measurement of Non Tariff Barriers
to Trade
Different measures of NTBs have been suggested
in the literature (Beghin & Bureau 2001; Disdier
et al. 2007)
•
•
•
•
Frequency and coverages type measure
Quantity impact measures
Price comparison measures
Price effect based on import demand
elasticities
Why MLRs of pesticides ? (1)
• Maximum Residue Limits or Maximum Level of
pesticides are established in most countries to
safeguard consumer health and to promote
Good Agricultural Practice (GAP) in the use of
insecticides, fungicides, herbicides and other
agricultural compounds.
• These MRLs vary from country to country
depending on the pesticides available, the crops
being treated and the way the pesticides are
used. Food exporters must comply with these
MRLs as a condition of market access.
Why MLR of pesticides ? (2)
• International harmonization of MLRs does not
exist at a global level.
• National authorities keep the sovereignity in
fixing these limits. Therefore these legal limits
can vary widely from a country to another.
• Standards and regulations differ across
importing countries and this heterogeneity will
cause standards and regulation to act as a
potential NTMs and thus impede trade.
Literature on MLR of pesticides (1)
• Otsuki and Wilson
and Sewadeh 2001
• Wilson and Otsuki
and Majumdar 2003
• Wilson and Otsuki
2004
• Wilson and Otsuki
2004
• Xion and Beghin 2010
• Level of Afaltoxins
• Level of tetracycline
• Level of chlorpyrifos
pesticides
• Level of Aflatoxins
• Level of Aflatoxins
Literature on MLR of pesticides (2)
• These studies share a common result that
more stringent food safety standards set
by developed countries tend to deter trade
supporting the view of standard as trade
barriers.
Hypothesis, model and results
Main actors
• Exporters are:
• EU
• Argentina
• Brazil
• China
• Chile
• New Zealand
• South Africa
They have been chosen on
the basis of market share
(first exporters)
• Importers are:
• USA
• Japan
• Mexico
• Korea
• Australia
• Russian Federation
• Canada
They have been chosen
based on the level of
imports and level of per
capita consumption
APPLES
Main
Exporters
Italy
Trade
Value
US$Mo
758.06
Trade
Quantity
‘000 tons
Main
Importers
798.30
Trade
Trade
Value
Quantity
US$Mo ‘000 tons
United
Kingdom
640.42
523.02
Germany
621.93
668.84
453.23
931.23
France
695.08
693.22
USA
651.29
663.47
China
512.65
1019.84
Russian
Federation
Chile
489.11
774.56
Netherlands
366.91
358.42
Netherlands
354.35
378.26
Spain
260.51
258.91
Belgium
268.11
342.05
Mexico
247.96
219.81
Belgium
210.88
227.63
USA
210.53
206.60
New
Zealand
265.30
322.49
South
Africa
212.66
334.34
Canada
178.70
180.49
Poland
173.29
449.73
Lithuania
91.93
172.38
PEARS
Main Exporters
pears
Netherlands
Argentina
Belgium
Italy
China
USA
South Africa
Spain
Chile
Rep. of Korea
Trade
Value
319.28
271.29
264.19
226.88
161.71
157.93
118.39
91.36
74.92
49.18
Trade
Quantity
320.01
Main Importers
pears
Trade
Value
Trade
Quantity
Russian
Federation
314.25
379.15
Germany
224.98
177.67
United Kingdom
159.23
130.01
USA
146.28
107.69
Netherlands
138.80
138.63
France
130.58
128.80
Italy
117.20
112.44
Brazil
98.05
137.44
Mexico
88.39
85.85
Canada
83.12
79.19
454.71
284.49
180.23
405.29
155.07
174.95
96.41
119.72
19.98
Multiple Markets
As many regulations as countries (1)
As many regulations as
countries (2)
• Countries’ regulation still differ a lot
• Most major exporters are developed countries
then they must
– Comply with their own domestic standards
– Meet different requirements across destination market
• We assume that a country which imposes tight
levels to its producers will be more capable to
export to countries with stringent regulations
Our Contribution
We consider:
• All pesticides regulated in the countries of
the sample (more or less 750).
• Differences in the regulation between
importing and exporting countries.
• Time variation of the MLR variable.
• Zero trade.
MLR INDEX (1)
• MLR = Weighted average (MLR_importer
– MLR_exporter). Weight is import of
pesticides
• This allows us to take into account the
differences between the legislation of the
importer and the exporter
The basic model
• We use a gravity model to estimate the impact of MLR
regulations on trade of apples, pear and related processed
products. The basic model has the following specification
k
lnX ijt
=0 +β1ln(GDPit)+β2 ln(GDPjt)+β3ln(POPit)+β4ln(POPjt)+β5 ln(DISTij )
+β6 ln(MLRkijt )+β7ln(Tariffijtk )+β8 Trasparencyit+β9Lang ij+β10Borderij+εkijt
:
Definitions of variables
Variables
Description
X
The trade volumes of product k from exporting countries j to importing
countries i in year t
GDPs
The Gross Domestic Product of importing countries i and exporting countries
j in year t
POP
The Popualation of importing and exporting countries
Dist
The distance between importing and exporting countries
MLR
The MLR is the difference in MLR applied to products k by importing i and
exporting j countries in year t
Transparency
The perception of corruption in importing countries i
Tariff
The tariff is an appled tariff rate in an advalorem term between countries i
and countries j
Lang
Dummy variable, coded 1 if the exporting country language is spoken in at
least one of the importing countries and 0 otherwise
Border
Dummy variable, coded 1 if the exporting country share the same border and
0 otherwise
MLR INDEX (2)
• A positve β suggests that MLR of
pesticides is trade-impeding: the lower the
tolerant level is, the less the bilateral trade
flows are
Databases
•
•
•
•
•
UNCOMTRADE (trade data at HS6 level)
WBDI (GDPs)
CEPII (Distance, applied tariffs)
FAS USDA (MLRs of pesticides in ppm)
TRASPARENCY.ORG (Index of
corruption)
Results (1)
OLS
LSDV
POISSON
NBR
ZIP
GDPIMP
0.492***
1.251***
0.809***
1.114***
0.633**
[0.105]
[0.343]
[0.195]
[0.297]
[0.248]
POPIMP
0.682***
-11.930***
-3.194***
-17.416***
-4.087*
[0.133]
[4.121]
[1.218]
[0.771]
[2.438]
GDPEXP
-0.624***
0.378
0.388
0.585
0.422
[0.087]
[0.336]
[0.411]
[0.513]
[0.428]
POPEXP
0.710***
18.443***
-9.166**
5.332*
-8.133***
[0.077]
[3.994]
[3.630]
[3.039]
[2.124]
DISTANCE
0.045
-0.584***
-0.443
-1.083***
-0.067
[0.115]
[0.115]
[0.328]
[0.108]
[0.214]
TARIFF
-0.130***
-0.156***
-0.274***
-0.291*
-0.270***
[0.028]
[0.037]
[0.074]
[0.149]
[0.071]
CORRUPTION
0.053*
0.01
0.020***
0.035**
0.026*
[0.030]
[0.026]
[0.006]
[0.015]
[0.013]
MLRINDEX
-1.045**
-1.348***
-1.093**
-1.545***
-0.879***
[0.437]
[0.358]
[0.510]
[0.268]
[0.282]
BORDER
0.961***
1.224***
-0.224
0.303
-0.151
[0.359]
[0.314]
[0.629]
[0.559]
[0.575]
LANGUAGE
1.051***
0.943***
1.025**
1.403***
0.657
[0.212]
[0.196]
[0.468]
[0.440]
[0.521]
COLONY
-2.081***
-0.379
-1.986***
0.117
-1.746***
[0.254]
[0.267]
[0.492]
[0.965]
[0.498]
Constant
-11.130***
-133.245
196.718**
176.859***
195.351**
[2.414]
[103.679]
[90.358]
[62.235]
[88.095]
Observations
1928
1928
5560
5560
5560
R-squared
0.24
0.72
0.36
0.05
0.35
Robust standard errors in brackets * significant at 10%; ** significant at 5%; *** significant at 1%
ZINBR
1.406***
[0.149]
-8.187***
[1.501]
0.309
[0.369]
12.503***
[4.032]
-0.373*
[0.221]
-0.199***
[0.043]
0.030***
[0.004]
-0.927***
[0.289]
0.034
[0.554]
0.999***
[0.174]
-0.753
[0.701]
-109.579
[67.133]
5560
0.16
Results (2)
MLR EU
MLR SAF
MLR NZ
MLR CHN
MLR CHL
MLR BRA
MLR ARG
POISSON
-2.299***
[0.695]
-0.408
[0.810]
-3.841**
[1.735]
0.275
[1.161]
-2.612***
[0.775]
-0.985***
[0.237]
-0.33
[1.242]
NBR
-1.678*
[0.912]
-2.496*
[1.337]
-9.753***
[1.064]
2.019*
[1.119]
-5.331***
[1.307]
-1.520**
[0.722]
-5.214***
[1.915]
ZIP
-2.299**
[0.900]
0.539
[1.032]
-2.848
[1.886]
-0.081
[0.580]
-1.838**
[0.828]
-0.911*
[0.507]
0.217
[1.227]
ZNB
-2.082***
[0.671]
-1.559**
[0.725]
-6.201***
[1.814]
1.944***
[0.416]
-1.567***
[0.509]
-1.322***
[0.405]
-0.613
[1.447]
Robustness
• MLR of importing countries in place of
MLR INDEX
• MLR time invaring
• Heckman two steps
• Hurdle double models
Comments (1)
• The first table presents the results with 6
different methods of estimations on the
entire sample.
• The MLR coefficient is negative and
significative, meaning that regulations on
pesticides may have a positive impact on
trade
Comments (2)
• MLR is interacted with 7 dummy variables
that respectively take the value 1 if
exporter are (i) EU, (ii) South Africa, (iii)
New Zealand, (iv) China, (v) Chile, (vi)
Brazil, (vii) Argentina
• Our analysis suggests that MLR is
impeding trade only for China, while in the
other cases MLR of pesticides enhance
trade
Conclusion (1)
• The results indicate a positive effect of MLRs
imposed by importing countries on pome
fruits exports
• A 1% increase in regulatory stringency –
tighter restrictions on MLRs - leads to an
increase of exports of pome fruits by 0.92%
This could be seen as confidence of
consumers
Conclusion (2)
• This result suggests that contrary to what is
commonly accepted, the stringency of
regulations in the area of hazardous
substances (like pesticides) is not tradeimpeding but rather trade-enhancing.
• It seems that it acts as a proof of trustability in
the safety product. Consumers are more
confident in products coming from developed
areas known for being more stringent than from
developing areas known for being laxer.
Conclusion (3)
• Costs of compliance are not a significant
problem because these countries must comply
with their own domestic regulations that are
already stringent.
• Then food safety standards may affect
negatively the competitiveness of developping
countries and positively that of developed
countries.
Develpoments
• Consider both MLR of pesticides for
importing and exporting countries
• Add other covariates in the gravity model
• Re-run the gravity model with a new data
on MLR of pesticides and replace the
missing value of the codex with 100 in
place of 75
• Check for endogeneity problems
Descargar

Do differences in levels of pesticides' MRLs affect trade