Abstract

In this paper, we introduce three indices of trade which are consequently applied to the case maximum residue level, MRLs of pesticide. The aggregation indices of NTM i.e Li and Beghin index of trade, Heterogeneous index of trade (HIT) and actual heterogeneous index of trade (AHIT) regulations to compute their protectionism relative to international standards. The application of indices to Maximum Residue Limit (MRL) affecting cottonseed and its by-product i.e crude cottonseed oil and refined cottonseed oil, using a science-based standards Codex Alimentarius. In this analysis, the main aim of the article is to show the response of the maximum residue limits (MRLs) adopted by 50 importing countries on exports of the cottonseed commodities for the year 2005 to 2014. In this study, we estimate the augmented gravity model using the ordinary least squares (OLS) and Poisson regression estimators. The results show that maximum residue limits (MRLs) have a trade increasing effect on exports of the cottonseed commodities.

Key Words

Cottonseed, Poisson, OLS, and MRLs

Introduction

Standards are of diverse types and have an unpredictable effect on trade. These are much more convenient and flexible and can be readily applied in times of emergency where the importing country needs to take quick action against certain country (Bergstrand, 1990). Moreover, increasing tariffs rates imports restrict imports indirectly by increasing cost and price of imported good. However, if an exporting country reacts to the policy of importing country through tax refund for stimulating exports or uses dumping sales strategies to reduce the effects of tariffs, then high tariffs rates will not bring the desired results (Krugman, 1979). 

Cotton is an important commercial cash crop in the world and a large source of fiber. Cotton is also used to produce more food for men and feed for the animal than fiber. Cotton seed is a byproduct of cotton, which further consist of the hull and kernel. The hull induces fiber and lint while, kernel carry protein, oil, carbohydrate, vitamins, and minerals. Food and Agriculture Organization (FAO) considers cottonseed oil as vegetable oil for food consumption. The top five cottonseed oil producers include China, India, Pakistan, the USA, and Uzbekistan.

Cottonseed oil is further a by-product of cottonseed and is an important source of edible oil. Cottonseed oil is pulled from cottonseed kernel. Cottonseed consists of 20% of oil and 38% of kernel, which further depends upon the seed quality. Cottonseed oil is also known as “Heart oil” which is among the most unsaturated edible oil. Meanwhile, the free fatty acid and oil depend on climatic condition in which cotton plant stands in the field. Quality of cottonseed oil depends upon time, place and season in which cotton plant stands in the field. Furthermore, the high-quality cottonseed oil is produced in dry weather and low-quality cottonseed oil is produced in wet weather or store in moisture. The crude cotton seed oil has good stability due to the presence of gossypol on cottonseed (Bambawale et al., 2004).

The Gossypol is removed from cottonseed oil during the refining process. It is a natural toxin that protects cotton plants from insect naturally harming insects and protects cotton plant (Kanoi, 2005). Non-refined cottonseed oil is used as a pesticide. Cottonseed oil has nonfood uses, too. It is used in cosmetic, laundry detergents and insecticide.  Cottonseed oil contains high concentration of vitamin E, fatty acid, an antioxidant that has benefit for human skins, moistening, anti-aging, and anti-inflammatory properties, certain fatty acids increase human skin penetrability. This allows human skin to better absorb other ingredients for better results. The crude oil contains a chemical known as aflatoxins which has strong flavor and odor. These are removed during the refining process to convert crude oil into excellent edible oil. Being healthy food, its demand has been growing with every passing day creating market expansion opportunities for cotton. 

The gospel present in cottonseed not only acts naturally against predators but also makes insects infertile by reducing sperm production in male insects. During cottonseed refining process, the gospel is removed to produce edible oil. The Codex Committee on 1967 declared that gospel is not a health hazard as it is removed during refining. 

Cottonseed oil with no gospel is yellow in color and full of vitamins-E, minerals and is used as cooking oil. As well as, used for manufacturing Vanaspati. The stored value of cottonseed oil is also good and is comparable with other edible oil (Bailey, 1948). Cottonseed oil is better than other edible oil as it lasts a long time in a relatively high temperature of food item due to its anti-oxidant content. For instance, chips and snacks may keep up a long shelf life. Cottonseed oil is a good option for preparing healthier food.

Cottonseed oil is used as a cooking oil and is also used as a home remedy for certain skin diseases. Like olive oil, cottonseed oil consists of polyunsaturated fatty acid which helps in lowering LDL (“bad” cholesterol) and increase HDL (“good” cholesterol). But it is also high in saturated fats, which has the opposite effect on cholesterol and increase the risk of heart disease.

Annually, cotton farmers harvest about 15 million bales of cotton, or about 7.3 billion pounds, six billion pounds of cottonseed and cottonseed meal are used in livestock feed, and more than 90 million gallons of cottonseed oil is used for human food products, such as salad oil dressing or margarine. The total production of cottonseed oil in the world was estimated to be 5.18 million metric tons during the period 2018 as compared to 5.12 million metric tons in 2014. Similarly, in the same period domestic consumption was 5.06 and 5.15. It is argued that an MRL  of less than 0.01 ppm in cottonseed covers residues of 2, 4-D. These residues in imported cotton do not pose any health hazard to any age group segment of the population.

Methods

The conceptual model used in the study is based on Haq, Meilke, and Cranfield (2013) and Haq and Meilke (2009, 2010). This study extended their model by considering non-tariff measures in the analysis. The model assumes that consumer in each importing country maximizes a constant elasticity of substitution utility function from the consumption of imported cotton products subject to his income constraint. The maximization problems assume that consumer has perfect information involved in the choice problem and yields demand for trade. Preferences are complete, reflexive and continuous.  The consumer is assumed to be price taker i.e. prices are ?xed and exogenous. Hence, the search for better prices, bargaining, and discount are ignored. Further prices are linear, and every unit of the cotton and textile product cost the same price. Hence, quantity discounts are assumed away. The maximization theory also assumes that goods are divisible. 

Cotton products are assumed to be differentiated. Substituting for trade cost in the demand function generates the following function.

lnM_in=?_i+?_j+?_1 lnD_ij+?_2 B_ij+?_3 T_ijn+?_4 ?MRL?_ij+?_5 ?PTA?_ij

+?_6 ?GSP?_ij+?_7 lnI_i+?_ij (1)

Jacks, Meissner, and Novy (2008), Haq et al. (2013) and Haq and Meilke (2010, 2009) have ignored the effect of standards such as MRLs in their analysis. However, this study contributes to the existing literature by explicitly considering non-tariff barriers in the analysis. We are unaware of a study that has considered both tariff and MRLs in the same study.




Data




Data required for present analysis pertain to the cottonseed oil for 50 countries with six-digit code of the Harmonized Commodity Description and Coding System (HS) disaggregation. The sample for the present study comprises the exporters for the time period of 2003 to 2014. The data for variables GDP, Population, US GDP deflator with the base year 2010 has been taken from World Development Indicators (WDI).  The data for common border and common official language (English) has been taken from Wikipedia. The Distance data is generated by http://www.distancefromto.net/countries.php. PTA variable incorporated in model is developed from Asian Regional Integrated center (ARIC) published by Asian Development Bank. Trade data is obtained from COMTRADE, a database on merchandize international trade maintained by the United Nation Statistical Division (UNSD). Moreover, the bilateral tariff rates have been accessed from TRAINS and World Integrated Trade Solution (WITS), a database maintained by UNCTAD for tariff and non-tariff barriers. For the sample group of countries various Generalized System of Preferences (GSP) schemes which are currently in operation according to notifications received from UNCTAD Member State, has been incorporated and taken from UNCTAD and various Official Journals of EU. 




Li and Beghin Index




The maximum amount of (residue level) of the importer “i” is for the good “j” and the harmful property; M_jn, let's give the international reflection of the rest of the same good and harmful substances. “i” is the total number of importers, the total number of products of “j”, the number of chemicals/pesticides used by the product “j” as n.  We offer the following indices,

S_ij=1/n (?_1^n?exp((M_(codx,jn-) M_ijn)/M_(codx,jn) ) )

Different levels of composition have special goals. The extent of the protection is to provide detailed information to the importers in goods. This aggregator is useful for economic studies of cross-sections (objects and countries). It is the SPS policy that uses gravity equations, as well as tariff subsidies for farmers in the financial sectors can assess the impact of these policies. These commodities in a country score are the structure blocks of the other two combined scores. By combining overall considered commodities by country, to create protectionism in these SPS rules. These combined country indexes give rank to rank according to our MRL protection system. These statistics are analyzing the total protection measure with tariffs and form financial assistance to investigate the policy composition of protectionism. Defining an importer’s MRL to be protectionist one if it is stricter than science-based international MRL and non-protectionist if the opposite is true. Compilation of various objects (materials) ensures the reliability of the indices. The "unexpected" protection MRL will also be known as MRL and other related products, such as the following three indicators in the country.

Heterogeneity Index of Trade (HIT) 




An introduction to the HIT list given by Rau et al. (2010) built up the HIT Index. The idea behind the HIT index compares the industry-based assumptions and country requirements for various needs and products that are suitable for agriculture business and process criteria. The HIT Index is designed to provide binary, configured, and measurable data extracted from appropriate requirements and documents in data collection efforts of NTM.

In light of the Gower index of similarity and dissimilarity, the condition for HIT index is as follows:

?                                               HIT?_jk=(?_(i=1)^n??W_ijk ?DS?_(ijk    ijk)^HIT ?)/(?_(i=1)^n?W_ijk )           1

where j represents import and k speak to the export of nation, and i indicates to the qualities or rather necessities took a look at. A few qualities or prerequisites could be easily compared with other, and is caught by the "W_ijn " .?DS?_ijn^HIT refers to a dissimilarity measure, which is defined by the following equation:

                                         ?DS?_ijk^HIT=|x_ij-x_in |/(max(x_i )-min(x_i ) )            2

where x indicates to the binary, orderd or quantitative data of the trademark or necessity, which the exporting and importing nation separately force. The HIT Index is a couple of collaborator countries. Therefore, it is based on the bilateral trade by comparing standards of importing and exporting countries. The estimation of the HIT list increments with the distinction in the standard.

The HIT index expects values somewhere in the range of 0 and 1. For (HIT) jn=0, there is no distinction in the prerequisites between the importing and exporting nation. For (HIT) jn=1 prerequisites are altogether different. Therefore, the HIT index gives data to the similarities in rule and laws in various countries and estimates the expenses incurred by exporters at the time of sale of products in foreign markets. It does not consider difference between trade and adaptability or trade costs laws. As noted above, the HIT index takes into consideration and compares different types of discriminations between countries and thus, the sphere of partner’s countries are alike or different. However, in order to interpret the HIT index, differences between countries result in expenditures from exports and affect the flow of international trade.




Actual Heterogeneity Index (AHI) 




The AHI Index created dependent on the presumptions that the contrasts among criteria and guidelines have dependably brought about contending expenses and exporters' trade costs. The MRL's fantasy example shows pesticides for pesticides to show that if the difference between exporters and importers is different from zero trade, it will not be necessary. Here is a picture of an exporting country Z, Y is a country importing pesticides MRL, A, B, C, and D. There is no difference between A and B, so (d) the similarity is zero. As the pesticide D, the exporting country (Z) is stricter than the importing country, so there is no need to arrange for MRLs placed in the exporting country. However, due to rising prices in the case of pesticides, MRL requirements of the export country are less strict than import.

The AHI appraisal in this section recognizes that the costs of reconciliation with the exporters are not differentiated. Hence, if you add the total value and effectively obedience to this difference, the resulting difference in the difference represents a fairly good difference, the actual cost implications are excessive. This becomes the measure of contradiction between national laws. A slight change is put forward to avoid this difference. If there are no legitimate expenses, these cases are identification and exemption. The first step to solving this problem is to learn not just a novelty but also a sign. This is an important indication that if the AHI index is calculated, countries will be honored at establishing a business relationship. This idea follows Achterbosch et al. (2009) that the exporters were stricter than the requirements for the exporters, and if needed, the heterogeneity index for pesticides was used.

The HIT-index formula is formulated in the equation (1), So AHI index is considered the average value of the actual heterogeneity as under.

?AHI?_jn=(?_(i=1)^k??W_ijn ?DS?_ijn^AHI ?)/(?_(i=i)^n?W_ijn )     3

Here the Wijn stand for weight of characteristics.

The AHI index is constructed using a modified structure formation, which is constructed so that HIT does not constitute a full value in the range, as in the formation of the contradiction of the index (comparable to equation 2).




The AHI Index has the Following Properties




The contradiction between MRL of the two countries indicates higher costs for exporters. This means that AHI only increases when higher export costs of exporters indicate a difference in control.

The index believes that the differences between the rules that do not require obedient value will not be included in the calculation at zero.

The values in the index indicate closer to zero, usually low interest and at least the costs of traders (support costs) for exporters. Values close to it indicate that the importers' needs are more stringent and that exporting countries should have higher costs to regulate their MRLs.

In the participating and exporting countries, the MRL situation corresponds to the pair.

The MRL contradiction between the two countries will be introduced only when higher exporters' costs are applied. This means that AHI only increases when higher export costs of exporters indicate a difference in control.

The difference between unnecessary costs and the regulatory costs required for the Index should not be included in the calculations to establish a zero level.

As a result, the index values in the index indicate zero, usually low base and possibly low trade costs (including marginal costs) among exporters. Values close to it indicate that the importers' needs are more stringent and that exporting countries should have higher costs to regulate their MRLs.

MRL case corresponding to pairs in import and export countries.

The calculation of the index with the participation of importers can be included in the calculation of the index.

The value of the AHI index can be combined with other regulatory and other subdirectories.

Properties of the Indices

In terms of design, the indices have the following characteristics. They are added to the MRL in relatively small amounts (%) and are completely free and go to unity because of common deviations. Units used for MRLs do not matter if the MRL and MRL codes are in the same units. The most stringent costs outside the code are growing. We regulate the control and regulatory body (the MRL in different countries, the same product and the same harmful as other things), and the link between the lower and upper.

First of all, we reduce MRL imports from the international MRL, because as Protocols only a portion of MRLs of more stringent importers will contribute to the Protocol. If the MRL is weaker than the international standard, it is anti-protection. 0.01 ppm (parts per million) 10 ppm or more. To change the index, depending on the number of different amounts of waste, we have international standards MRL. Secondly, for a low (ie, strict) standard export is more difficult. For example, some importers, with the exception of MRL, are international. But a strict MRL can harden. Taking the status of Protestants, which includes the MRL role. So we see that the more stringent MRLs look more burdensome.

Finally, if the indices are the same j, n, all other things are equal, different countries have a stable speed (M (ijn), and if the country's level becomes more strict, then in an emergency it will improve in an emergency situation 0 'Fight Night' (MRL) and the aforementioned bike (zero MRL) and enter -2.72, Score points to 1 security policy (MRL in MRL code) Points above 1 indicate "protectionism", not MRL, which is strictly greater, than on codex, and estimates below 1 are "anti-spacing", codex Indicates that this is the key.

Results and Discussion

As discussed in the previous section, MRLs is measured using Li-and Beghin, Heterogeneous Index of Trade and Absolute Heterogeneous Index of Trade. The effect of these indices on cottonseed, crude oil and refined oil are estimated using Poisson and OLS techniques. Estimated results in case of cottonseed are presented in tables 4.5 to 4.7, for crude cottonseed oil in tables 4.8 to 4.10, and for refined cottonseed oil in tables 4.11 to 4.13. It is important to mention that crude (151221) and refined (151229) cottonseed oil  and cottonseed (120720) are disaggregated at 6-digit level of disaggregation.  The effect of MRLs on the trade of crude and refine cottonseed oil (1512), collectively representing both refined and crude oil, is also estimated and results are presented in table 4.15 to 4.16.

The tables for cottonseed show that the total number of observations are 2,062. However, only 710 observations are left for the analysis when zero trade-flow are not omitted. Hence, two-third of the observations are zero trade flows. Similarly, in the case of crude seed oil three-quarter of the observations are zero-trade flow while it is two-third for the refine oilseed. Hence, on average only one-third of the observations are used in the OLS regression analysis as with logarithm specification, zero observations become missing and that is the point for using Poisson regression.

The Pseudo R-squared for the three commodities is very high and range from 0.514 for refining seed oil (tables 11 to 13) to 0.618 for cottonseed (tables 5 to 7). The Pseudo R-squared for the overall cottonseed oil  (tables 14 to 16) also falls in the same range. These are called “pseudo” because they are similar to R-squared as they also fall between 0 to 1 interval but these are not interpreted as R-squared. In the case of R-squared it ranges from 0.563 for refined seed oil to 0.666 for crude seed oil. In case of regression analysis, R-squared is an important indicator for at least three reasons. First, it explains the variability of the dependent variable from its mean that is the proportion of the total variation as unexplained by the model. Second the denominator of the R-squared formula (that is R^2=1-(?_(i=1)^n?(y_i-(y_i ) ? )^2 )/(?_(i=1)^n?(y_i-(y_i ) ? )^2 )) represents sum of the squared errors of the null model. The null model shows the predicting of the endogenous variable without exogenous variables while the numerator shows the sum of squared errors of the fitted model. Hence, the ratio clearly shows improvement in the prediction power of the model due to exogenous variables. Third, R-squared can also be determined as the square of the correlation between the predicted and actual values of the model. Hence, High Square is not only indicative of the good explanatory power but also strong predictive power of a model. The F-statistics and Wald Chi-squared show that all the models are statistically significant that is the effect of all the exogenous variables excluding the intercept on the dependent variable is statistically significant. All the fixed effects controlling for the importing and exporting countries are statistically significant implying that ignoring these effects would lead to biased estimates. However, the year fixed effects are statistically significant 20 times out of 27.

The effect of distance on trade is assumed to be negative while common border, Preferential Trade agreement (PTA) and General System of Trade Preferences (GSP) is supposed to have positive effect on trade. The estimated results show that the effect of distance on trade of all the commodities is negative and statistically significant. The marginal effect (that is elasticities) of distance on trade estimated using Poisson is consistently higher as compared to the elasticities estimated using OLS. Elasticities are higher for crude seed oil as compared to others. The effect of common border on seed cotton and crude seed oil are consistently positive and statistically significant while it does not affect refined seed oil trade. Similarly, the effect of PTAs on trade of all the three commodities is positive when it is statistically significant. In the case of PTA, out of the 36 estimated parameters (tables 4.5 to 4.16) only seven have statistically insignificant effect on trade. However, the effect of PTA estimated using Poisson is much smaller than those estimated using OLS. The effect of GSP on trade is predominantly statistically insignificant. However, its effect is positive whenever it is statistically significant.

Tariff is an important determinant of crude seed oil trade only. Its effect is negative and statistically significant.  Its elasticity estimated using Poisson is highly elastic as compared to the elasticities estimated using OLS (Tables 8 to 10). This is the first evidence of its kind of the effect of the tariff on crude oil trade. Poisson estimates show that, on average, a 10 percent increase in tariff reduces crude oil trade by about 25 percent, keeping other variables constant while the same effect estimated using OLS is about 14 percent. When commodities are aggregated to four-digit that is cottonseed oil (tables 4.14 to 4.16), the effect tariff on trade fades away as none of the parameters of simple average tariff is statistically significant. 

This gets us to the effect of MRLs on the trade of the selected commodities. However, before discussing its result it is important to understand the context of the cottonseed crude oil trade. Since it is not very widely traded and based on its value it is ranked at about 4500 in the list of commodities traded and considered to the complex commodity to be traded. The complexity is due to the biology of the cottonseed. 

Seed account for two-thirds of the cottonseed that is both cotton boll, seed, and lint. There could be physical barriers to the utilization of seed because of a chemical tetraploidy, largely available in the crude oil. The other is that cottonseed could deteriorate due to non-availability of storage to keep the seed cool and dry and stop degradation such as production of fatty acid, odor, and a darkening of the seed (Gregory et al. 1995).

While there are MRL standards for crude oil but countries and especially the EU do not specify these standards for processed products like crude oils. In such a situation when MRL is not defined for a processed food product, then the upper limit of MRLs is set equal to the MRL of raw product in this case cottonseed. The concentration of the product during the refining process is also considered in the MRL determination. Since concentration process could be carried chemically as well as physically, therefore, the standard also considers the type of processing depending on whether the processing eliminates a pesticide sprayed on the plant, the source of food that is cottonseed.  

Hence, the allowable bandwidths of MRLs in cottonseed vary according to the chemical nature of the pesticides and the oil content. If a pesticide is highly soluble in fat or difficult to be eliminated during the primary extraction process, then MRL is determined by multiplying the seed MRL with concentration factor as discussed above. Pesticides might also be solvable in water or fat or both. The bottom line is that untraceable traces of chemicals might exist in cottonseeds, and concentration during initial processing would lead its detection in the crude oil. Hence, MRL standards are typically set higher in cottonseed, followed by crude and refined oil to protect human health. 

MRLs do not effect cottonseed trade. Results compiled in tables 4.5 to 4.7 show that the effect of MRLs on cottonseed is more prominent. Cottonseed is the source of crude and refined oil. However, it also gets all the pesticide sprayed on cotton plant and has the highest potential to absorb these. MRLs measured using Heterogeneous Index (HI) and Actual Heterogeneous Index (AHI) have a significant effect on crude cottonseed oil trade. The heterogeneity index of trade (HIT) of NTMs  by Rau et al. (2010) is binary in nature and measures the dissimilarity of NTMs of importing and exporting countries. However, countries could be dissimilar or otherwise in stringency of regulatory environment. In such case the effect of HI could be more intense. These effects are primarily inelastic except the marginal effect estimated using OLS for AHI which is highly elastic (–6.989). The AHI also shows statistically significant effect of MRLs on cottonseed trade, though its effect is smaller than those estimated with HI. The effect of MRLs on crude oil trade is statistically insignificant for all the indices and estimation procedure. In the case of refined cottonseed oil trade, the effect of MRLs on trade is statistically significant for one-third of the cases. In refined oil case, both the Li and Beghin and HI indices show statistically significant effect of MRLs on refined cottonseed oil. Finally, the overall effect of MRLs on cottonseed oil is statically significant as presented in tables 4.14 to 4.16. These tables consistently show statistically significant and negative effect of MRLs on trade. 

Conclusion

Irrespective of the measure, MRLs have a statistically significant negative effect on cottonseed and refine cottonseed oil. Since none of the measures of MRLs is perfect, therefore, if even one index shows a statistically significant negative effect on trade, then one can argue about the stringency of NTMs. The analysis showed that measure of MRLs and estimation technique affects not only the size of the estimate of MRL measure but also it’s statistically significant. Poisson uses more observations as compared to log-linearized OLS regression and hence produces statistically significant estimates. 

Table 1

OLS and Poisson estimates of the effect of tariff and MRLs on the trade of cottonseed

> > > > > > > > > > > > > > > > > > > > > > > > > > > > >

width="211" nowrap="" colspan="3"> LI and Bheghin width="206" nowrap="" colspan="3"> HI width="206" nowrap="" colspan="3"> AHI

Variables width="70"> Poisson width="66"> Marginal width="75" rowspan="2"> OLS Estimates (Logarithm) width="70"> Poisson width="61"> Marginal width="75" rowspan="2"> OLS Estimates (Logarithm) width="66"> Poisson width="66"> Marginal width="75" rowspan="2"> OLS Estimates (Logarithm)

Estimates width="66"> Effects width="70"> Estimates width="61"> Effects width="66"> Estimates width="66"> Effects

Distance width="70"> -0.000*** width="66"> -1.776*** width="75"> -0.977*** width="70"> -0.000*** width="61"> -1.777*** width="75"> -0.980*** width="66"> -0.000*** width="66"> -1.776*** width="75"> -0.975***

0 width="66"> -0.373 width="75"> -0.182 width="70"> 0 width="61"> -0.375 width="75"> -0.182 width="66"> 0 width="66"> -0.376 width="75"> -0.182

Common Border width="70"> 1.886*** width="66"> 0.169*** width="75"> 0.934** width="70"> 1.885*** width="61"> 0.169** width="75"> 0.920* width="66"> 1.888** width="66"> 0.169*** width="75"> 0.949**

-0.366 width="66"> -0.032 width="75"> -0.501 width="70"> -0.364 width="61"> -0.032 width="75"> -0.499 width="66"> -0.363 width="66"> -0.032 width="75"> -0.497

Partner Simple Average Tariff width="70"> 0.051 width="66"> 0.248 width="75"> 0.013 width="70"> 0.051 width="61"> 0.247 width="75"> 0.019 width="66"> 0.036 width="66"> 0.177 width="75"> 0.016

-0.044 width="66"> -0.215 width="75"> -0.032 width="70"> -0.039 width="61"> -0.184 width="75"> -0.032 width="66"> -0.031 width="66"> -0.151 width="75"> -0.031

MRL Index width="70"> 0.007 width="66"> 0.012 width="75"> 0.012 width="70"> -2.048** width="61"> -0.235** width="75"> -2.708 width="66"> -8.256*** width="66"> -0.533*** width="75"> -6.989**

-0.013 width="66"> -0.024 width="75"> -0.023 width="70"> -1 width="61"> -0.115 width="75"> -1.514 width="66"> -2.113 width="66"> -0.136 width="75"> -2.91

Preferential Trade Agreements width="70"> 0.793 width="66"> 0.163* width="75"> 0.842* width="70"> 0.793*** width="61"> 0.163* width="75"> 0.869* width="66"> 0.792* width="66"> 0.162* width="75"> 0.849*

-0.332 width="66"> -0.068 width="75"> -0.474 width="70"> -0.33 width="61"> -0.067 width="75"> -0.474 width="66"> -0.324 width="66"> -0.067 width="75"> -0.472

Generalized System Preferences width="70"> -0.835 width="66"> -0.133 width="75"> 0.329 width="70"> -0.823 width="61"> -0.131 width="75"> 0.371 width="66"> -0.83 width="66"> -0.132 width="75"> 0.367

-0.866 width="66"> -0.138 width="75"> -0.754 width="70"> -0.858 width="61"> -0.137 width="75"> -0.752 width="66"> -0.847 width="66"> -0.135 width="75"> -0.752

Real PCI of  Reporting country width="70"> -903088.5 width="66"> -0.155 width="75"> -3.395 width="70"> -737474.4 width="61"> -0.127 width="75"> -3.504 width="66"> -1386909 width="66"> -0.238 width="75"> -3.234

-1313428 width="66"> -0.22 width="75"> -4.131 width="70"> -1274897 width="61"> -0.219 width="75"> -4.15 width="66"> -1360445 width="66"> -0.234 width="75"> -4.114

Constant width="70"> 0.686 width="66"> width="75"> -61.79 width="70"> 0.727 width="61"> width="75"> -63.799 width="66"> 0.807 width="66"> width="75"> -58.67

-0.993 width="66">   width="75"> -78.549 width="70"> -0.994 width="61">   width="75"> -78.877 width="66"> -0.981 width="66">   width="75"> -78.21

Fixed Effects width="70"> width="66"> width="75"> width="70"> width="61"> width="75"> width="66"> width="66"> width="75">

Partner Country width="70"> 1155.400**** width="66"> ---- width="75"> 16.640*** width="70"> 1112.630**** width="61"> ---- width="75"> 16.360*** width="66"> 1201.900**** width="66"> ---- width="75"> 17.080***

Reporting Country width="70"> 3090.580*** width="66"> ---- width="75">       15.120*** width="70"> 3073.570*** width="61"> ---- width="75"> 15.370*** width="66"> 3191.910*** width="66"> ---- width="75"> 17.290***

Years width="70"> 2.130*** width="66"> ---- width="75"> 1.52 width="70"> 4.68 width="61"> ---- width="75"> 2.000** width="66"> 4.800*** width="66"> ---- width="75"> 1.13

Summary Statistics width="70"> width="66"> width="75"> width="70"> width="61"> width="75"> width="66"> width="66"> width="75">

Number of Observations width="70"> 2,062 width="66"> 2,062 width="75"> 710 width="70"> 2,062 width="61"> 2,062 width="75"> 710 width="66"> 2,062 width="66"> 2,062 width="75"> 710

Wald chi-square width="70"> 8886.040*** width="66"> ---- width="75"> ---- width="70"> 9069.860*** width="61"> ---- width="75"> ---- width="66"> 9694.090*** width="66"> ---- width="75"> ----

Pseudo R-squared width="70"> 0.611 width="66"> ---- width="75"> ---- width="70"> 0.613 width="61"> ---- width="75"> ---- width="66"> 0.618 width="66"> ---- width="75"> ----

F-statistics width="70"> ---- width="66"> ---- width="75"> . width="70"> ---- width="61"> ---- width="75"> . width="66"> ---- width="66"> ---- width="75"> .

R-squared width="70"> ---- width="66"> ---- width="75"> 0.57 width="70"> ---- width="61"> ---- width="75"> 0.579 width="66"> ---- width="66"> ---- width="75"> 0.578

RMSE width="70"> ---- width="66"> ---- width="75"> 2.655 width="70"> ---- width="61"> ---- width="75"> 2.65 width="66"> ---- width="66"> ---- width="75"> 2.647

Table 2: OLS and Poisson estimates of the effect of tariff and MRLs on trade of crude cottonseed oil

> > > > > > > > > > > > > > > > > > > > > > > > > > > > >

width="217" nowrap="" colspan="3"> LI and Bheghin width="207" nowrap="" colspan="3"> HI width="210" nowrap="" colspan="3"> AHI

Variables width="71"> Poisson width="71"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm) width="66"> Poisson width="66"> Marginal width="75" rowspan="2"> OLS Estimates (Logarithm) width="61"> Poisson width="66"> Marginal width="82" rowspan="2"> OLS Estimates (Logarithm)

Estimates width="71"> Effects width="66"> Estimates width="66"> Effects width="61"> Estimates width="66"> Effects

Distance width="71"> -0.000*** width="71"> -3.542*** width="76"> -o.515 width="66"> -0.000*** width="66"> -3.672*** width="75"> -0.492 width="61"> -0.000*** width="66"> -3.613*** width="82"> -0.496

0 width="71"> -0.987 width="76"> -0.447 width="66"> 0 width="66"> -0.987 width="75"> -0.447 width="61"> 0 width="66"> -0.986 width="82"> -0.447

Common Border width="71"> -0.277 width="71"> -0.045 width="76"> 0.866 width="66"> -0.303 width="66"> -0.048 width="75"> 0.866 width="61"> -0.284 width="66"> -0.045 width="82"> 0.866

-0.624 width="71"> -0.101 width="76"> -0.602 width="66"> -0.634 width="66"> -0.102 width="75"> -0.597 width="61"> -0.636 width="66"> -0.103 width="82"> -0.598

Partner Simple Average Tariff width="71"> -0.266*** width="71"> -2.338*** width="76"> -0.142*** width="66"> -0.287*** width="66"> -2.526*** width="75"> -0132*** width="61"> -0.281*** width="66"> -2.473*** width="82"> -0.135***

-0.052 width="71"> -0.457 width="76"> -0.037 width="66"> -0.05 width="66"> (0.0.439) width="75"> -0.037 width="61"> -0.049 width="66"> -0.435 width="82"> -0.037

MRL Index width="71"> -0.548 width="71"> -0.398 width="76"> 0.713 width="66"> -1.602 width="66"> -0.293 width="75"> 1.0321 width="61"> -1.172 width="66"> -0.154 width="82"> 1.074

-0.722 width="71"> -0.524 width="76"> -0.578 width="66"> -1.148 width="66"> -0.211 width="75"> -1.321 width="61"> -1.374 width="66"> -0.18 width="82"> -1.868

Preferential Trade Agreements width="71"> 3.945*** width="71"> 1.449*** width="76"> 1.68 width="66"> 3.941*** width="66"> 1.447*** width="75"> 1.694 width="61"> 3.934*** width="66"> 1.445*** width="82"> 1.668

-1.079 width="71"> -0.397 width="76"> -1.475 width="66"> -1.071 width="66"> -0.394 width="75"> -1.475 width="61"> -1.084 width="66"> -0.398 width="82"> -1.472

GSP width="71"> 0.189 width="71"> 0.012 width="76"> -3.961* width="66"> 0.459 width="66"> 0.028 width="75"> -3.432* width="61"> 0.414 width="66"> 0.025 width="82"> -3.534

-1.666 width="71"> -0.104 width="76"> -1.526 width="66"> -1.664 width="66"> -0.104 width="75"> -1.528 width="61"> -1.664 width="66"> -0.104 width="82"> -1.498

Real PCI of  Reporting country width="71"> 2831214 width="71"> 0.539 width="76"> -2.511 width="66"> 3146054 width="66"> 0.599 width="75"> -2.033 width="61"> 3050966*** width="66"> 0.582*** width="82"> -2.168

-603124.1 width="71"> -0.115 width="76"> -4.595 width="66"> -716068 width="66"> -0.136 width="75"> 4.653 width="61"> -621534.1 width="66"> -0.118 width="82"> -4.638

Constant width="71"> -5.111** width="71"> width="76"> -52.205 width="66"> -5.206 width="66"> width="75"> -42.623 width="61"> -5.204 width="66"> width="82"> -45.192

-1.835 width="71">   width="76"> -87.709 width="66"> -1.839 width="66">   width="75"> -88.769 width="61"> -1.849 width="66">   width="82"> -88.507

Fixed Effects width="71"> width="71"> width="76"> width="66"> width="66"> width="75"> width="61"> width="66"> width="82">

Partner Country width="71"> 1290.500**** width="71"> --- width="76"> 3.610*** width="66"> 1579.92**** width="66"> ---- width="75"> 3.590*** width="61"> 1595.070**** width="66"> ---- width="82"> 3.58***

Reporting Country width="71"> 402.500*** width="71"> ---- width="76"> 14.210*** width="66"> 421.670*** width="66"> ---- width="75"> 17.030*** width="61"> 409.540*** width="66"> ---- width="82"> 18.870***

Years width="71"> 141.890*** width="71"> ---- width="76"> 2.670** width="66"> 144.610*** width="66"> ---- width="75"> 1.59 width="61"> 147.310*** width="66"> ---- width="82"> 2.230*

Summary Statistics width="71"> width="71"> width="76"> width="66"> width="66"> width="75"> width="61"> width="66"> width="82">

Number of Observations width="71"> 1,226 width="71"> 1,226 width="76"> 327 width="66"> 1,266 width="66"> 1,266 width="75"> 327 width="61"> 1,266 width="66"> 1,266 width="82"> 327

Wald chi-square width="71"> 7028.310*** width="71"> ---- width="76"> ---- width="66"> 7795.14 width="66"> ---- width="75"> ---- width="61"> 7551.120*** width="66"> ---- width="82"> ----

Pseudo R-squared width="71"> 0.556 width="71"> ---- width="76"> ---- width="66"> 0.557 width="66"> ---- width="75"> ---- width="61"> 0.555 width="66"> ---- width="82"> ----

F-statistics width="71"> ---- width="71"> ---- width="76"> . width="66"> ---- width="66"> ---- width="75"> . width="61"> ---- width="66"> ---- width="82"> .

R-squared width="71"> ---- width="71"> ---- width="76"> 0.666 width="66"> ---- width="66"> ---- width="75"> 0.666 width="61"> ---- width="66"> ---- width="82"> 0.665

RMSE width="71"> ---- width="71"> ---- width="76"> 2.166 width="66"> ---- width="66"> ---- width="75"> 2.17 width="61"> ---- width="66"> ---- width="82"> 2.171

Table 3: OLS and Poisson estimates of the effect of tariff and MRLs on trade of refined cottonseed oil

> > > > > > > > > > > > > > > > > > > > > > > > > > > > >

width="217" nowrap="" colspan="3"> LI and Bheghin width="208" nowrap="" colspan="3"> HI width="211" nowrap="" colspan="3"> AHI

Variables width="71"> Poisson width="71"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm) width="66"> Poisson width="66"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm) width="66"> Poisson width="66"> Marginal width="79" rowspan="2"> OLS Estimates (Logarithm)

Estimates width="71"> Effects width="66"> Estimates width="66"> Effects width="66"> Estimates width="66"> Effects

Distance width="71"> -0.000*** width="71"> -1.211*** width="76"> -1.050*** width="66"> -0.000*** width="66"> -1.204*** width="76"> -1.048*** width="66"> -0.000*** width="66"> -1.203*** width="79"> -1.049***

0 width="71"> -0.323 width="76"> -0.149 width="66"> 0 width="66"> -0.327 width="76"> -0.149 width="66"> 0 width="66"> -0.327 width="79"> -0.149

Common Border width="71"> 2.763*** width="71"> 0.231*** width="76"> 0.253 width="66"> 2.755*** width="66"> 0.231*** width="76"> 0.264 width="66"> 2.756*** width="66"> 0.231*** width="79"> 0.263

-0.537 width="71"> -0.045 width="76"> -0.359 width="66"> -0.538 width="66"> -0.045 width="76"> -0.36 width="66"> -0.538 width="66"> -0.045 width="79"> -0.36

Partner Simple Average Tariff width="71"> -0.0383 width="71"> -0.325 width="76"> -0.004 width="66"> -0.044 width="66"> -0.376 width="76"> -0.006 width="66"> -0.0418 width="66"> -0.355 width="79"> -0.005

-0.032 width="71"> -0.271 width="76"> -0.022 width="66"> -0.032 width="66"> -0.275 width="76"> -0.022 width="66"> -0.032 width="66"> -0.275 width="79"> -0.0222

MRL Index width="71"> -0.977* width="71"> -0.723* width="76"> -0.376 width="66"> -0.627 width="66"> -0.118* width="76"> -0.144 width="66"> -0.592 width="66"> -0.081 width="79"> -0.194

-0.397 width="71"> -0.294 width="76"> -0.319 width="66"> -0.369 width="66"> -0.069 width="76"> -0.414 width="66"> -0.3694 width="66"> -0.05 width="79"> -0.438

Preferential Trade Agreements width="71"> 2.528*** width="71"> 0.626*** width="76"> -0.316 width="66"> 2.525*** width="66"> 0.625*** width="76"> -0.319 width="66"> 2.525*** width="66"> 0.625*** width="79"> -0.32

-0.829 width="71"> -0.205 width="76"> -0.398 width="66"> -0.83 width="66"> -0.206 width="76"> -0.399 width="66"> -0.83 width="66"> -0.205 width="79"> -0.399

GSP width="71"> 0.066 width="71"> 0.007 width="76"> 0.025 width="66"> 0.058 width="66"> 0.006 width="76"> 0.004 width="66"> 0.449 width="66"> 0.005 width="79"> 0.0006

-0.521 width="71"> -0.062 width="76"> -0.525 width="66"> -0.626 width="66"> -0.074 width="76"> -0.527 width="66"> -0.625 width="66"> -0.075 width="79"> -0.527

Real PCI of  Reporting country width="71"> -974.501 width="71"> -0.659 width="76"> -5.544** width="66"> -707.762 width="66"> -0.121 width="76"> -5.352 width="66"> -774.723 width="66"> -0.132 width="79"> -5.335*

width="71"> -849.585 width="71"> -0.144 width="76"> -2.743 width="66"> -772.311 width="66"> -0.131 width="76"> -2.743 width="66"> -796.61 width="66"> -0.074 width="79"> -2.738

Constant width="71"> -4.476*** width="71"> width="76"> -104.067 width="66"> -4.63 width="66"> width="76"> -100.612 width="66"> -4.629*** width="66"> width="79"> -100.287*

-0.885 width="71">   width="76"> -52.222 width="66"> -0.903 width="66">   width="76"> -52.227 width="66"> -0.902 width="66">   width="79"> -52.134

Fixed Effects width="71"> width="71"> width="76"> width="66"> width="66"> width="76"> width="66"> width="66"> width="79">

Partner Country width="71"> 5711.12**** width="71"> --- width="76"> 14.360*** width="66"> 5633.06**** width="66"> ---- width="76"> 15.380*** width="66"> 5822.760**** width="66"> ---- width="79"> 15.340***

Reporting Country width="71"> 2325.610*** width="71"> ---- width="76"> 11.88*** width="66"> 2363.820*** width="66"> ---- width="76"> 12.070*** width="66"> 2359.95*** width="66"> ---- width="79"> 12.100***

Years width="71"> 30.510*** width="71"> ---- width="76"> 0.95 width="66"> 20.65*** width="66"> ---- width="76"> 0.97 width="66"> 19.800* width="66"> ---- width="79"> 0.98

Summary Statistics width="71"> width="71"> width="76"> width="66"> width="66"> width="76"> width="66"> width="66"> width="79">

Number of Observations width="71"> 3546 width="71"> 3546 width="76"> 1199 width="66"> 3,546 width="66"> 3,546 width="76"> 1,199 width="66"> 3,546 width="66"> 3,546 width="79"> 1,199

Wald chi-square width="71"> 15197.670*** width="71"> ---- width="76"> ---- width="66"> 15704.88 width="66"> ---- width="76"> ---- width="66"> 15961.83*** width="66"> ---- width="79"> ----

Pseudo R-squared width="71"> 0.514 width="71"> ---- width="76"> ---- width="66"> 0.514 width="66"> ---- width="76"> ---- width="66"> 0.514 width="66"> ---- width="79"> ----

F-statistics width="71"> ---- width="71"> ---- width="76"> . width="66"> ---- width="66"> ---- width="76"> . width="66"> ---- width="66"> ---- width="79"> .

R-squared width="71"> ---- width="71"> ---- width="76"> 0.563 width="66"> ---- width="66"> ---- width="76"> 0.563 width="66"> ---- width="66"> ---- width="79"> 0.562

RMSE width="71"> ---- width="71"> ---- width="76"> 2.444 width="66"> ---- width="66"> ---- width="76"> 2.445 width="66"> ---- width="66"> ---- width="79"> 2.445

 

 

 

 

Table 4: OLS and Poisson estimates of the effect of tariff and MRLs on trade of cottonseed oil

> > > > > > > > > > > > > > > > > > > > > > > > > > > > >

width="209" nowrap="" colspan="3"> LI and Bheghin width="213" nowrap="" colspan="3"> HI width="213" nowrap="" colspan="3"> AHI

Variables width="71"> Poisson width="62"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm) width="66"> Poisson width="71"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm) width="71"> Poisson width="66"> Marginal width="76" rowspan="2"> OLS Estimates (Logarithm)

Estimates width="62"> Effects width="66"> Estimates width="71"> Effects width="71"> Estimates width="66"> Effects

Distance width="71"> -0.000*** width="62"> -1.814*** width="76"> -1.358*** width="66"> -0.000*** width="71"> -1.814*** width="76"> -1.358*** width="71"> -0.000*** width="66"> -1.814*** width="76"> -1.359***

0 width="62"> -0.156 width="76"> -0.075 width="66"> 0 width="71"> -0.154 width="76"> -0.075 width="71"> 0 width="66"> -0.153 width="76"> -0.075

Common Border width="71"> 1.693*** width="62"> 0.937*** width="76"> 1.848*** width="66"> 1.692*** width="71"> 0.094*** width="76"> 1.848*** width="71"> 1.692*** width="66"> 0.094*** width="76"> 1.847***

-0.109 width="62"> -0.006 width="76"> -0.165 width="66"> -0.111 width="71"> -0.006 width="76"> -0.165 width="71"> -0.111 width="66"> -0.006 width="76"> -0.165

Partner Simple Average Tariff width="71"> -0.01 width="62"> -0.086 width="76"> 0.001 width="66"> -0.011 width="71"> -0.102 width="76"> 0.001 width="71"> -0.011 width="66"> -0.1 width="76"> 0.001

-0.017 width="62"> -0.146 width="76"> -0.012 width="66"> -0.017 width="71"> -0.144 width="76"> -0.012 width="71"> -0.017 width="66"> -0.144 width="76"> -0.012

MRLs Index width="71"> 0.316* width="62"> 0.245* width="76"> 0.017 width="66"> 0.226* width="71"> -0.047* width="76"> -0.198 width="71"> -0.247* width="66"> -0.038* width="76"> -0.218

-0.181 width="62"> -0.141 width="76"> -0.159 width="66"> -0.105 width="71"> -0.022 width="76"> -0.127 width="71"> -0.109 width="66"> -0.017 width="76"> (0.128)*

Preferential Trade Agreements width="71"> 1.585*** width="62"> 0.319*** width="76"> 0.319*** width="66"> 1.585*** width="71"> 0.319*** width="76"> 0.319*** width="71"> 1.585*** width="66"> 0.319*** width="76"> 0.319***

-0.177 width="62"> -0.036 width="76"> -0.144 width="66"> -0.177 width="71"> -0.036 width="76"> -0.144 width="71"> -0.177 width="66"> -0.036 width="76"> -0.144

GSP width="71"> -0.473 width="62"> -0.069 width="76"> 0.644 width="66"> -0.449 width="71"> -0.065 width="76"> 0.646** width="71"> -0.453 width="66"> -0.066 width="76"> 0.646**

-0.33 width="62"> -0.048 width="76"> -0.262 width="66"> -0.326 width="71"> -0.048 width="76"> -0.262 width="71"> -0.326 width="66"> -0.048 width="76"> -0.262

Real PCI of  Reporting country width="71"> 291.1 width="62"> 0.057 width="76"> 0.312 width="66"> 301.3 width="71"> 0.058 width="76"> 0.297 width="71"> 302.2 width="66"> 0.058 width="76"> 0.299

width="71"> -202.4 width="62"> -0.039 width="76"> -0.922 width="66"> -200.7 width="71"> -0.039 width="76"> -0.923 width="71"> -200.7 width="66"> -0.039 width="76"> -0.923

Constant width="71"> -2.728*** width="62"> width="76"> -4.455* width="66"> -2.490*** width="71"> width="76"> 11.928 width="71"> -2.470*** width="66"> width="76"> 11.98

(0..312) width="62">   width="76"> -2.608 width="66"> -0.281 width="71">   width="76"> -17.589 width="71"> -0.281 width="66">   width="76"> -17.591

Fixed Effects width="71"> width="62"> width="76"> width="66"> width="71"> width="76"> width="71"> width="66"> width="76">

Partner Country width="71"> 4011.230**** width="62"> ---- width="76"> 131.220*** width="66"> 5054.170**** width="71"> ---- width="76"> 145.830*** width="71"> 5055.280**** width="66"> ---- width="76"> 145.760***

Reporting Country width="71"> 4065.210*** width="62"> ---- width="76"> 36.000*** width="66"> 4015.810*** width="71"> ---- width="76"> 36.030*** width="71"> 4009.170*** width="66"> ---- width="76"> 36.040***

Years width="71"> 33.530*** width="62"> ---- width="76"> 3.140** width="66"> 48.440*** width="71"> ---- width="76"> 5.320** width="71"> 47.720*** width="66"> ---- width="76"> 5.310**

Summary Statistics width="71"> width="62"> width="76"> width="66"> width="71"> width="76"> width="71"> width="66"> width="76">

Number of Observations width="71"> 9,729 width="62"> ---- width="76"> 5,296 width="66"> 9,729 width="71"> 9,729 width="76"> 5,296 width="71"> 9,729 width="66"> 9,729 width="76"> 5,296

Wald chi-square width="71"> 11315.790*** width="62"> ---- width="76"> ---- width="66"> 11083.000*** width="71"> ---- width="76"> ---- width="71"> 11079.090*** width="66"> ---- width="76"> ----

Pseudo R-squared width="71"> 0.568 width="62"> ---- width="76"> ---- width="66"> 0.568 width="71"> ---- width="76"> ---- width="71"> 0.568 width="66"> ---- width="76"> ----

F-statistics width="71"> ---- width="62"> ---- width="76"> . width="66"> ---- width="71"> ---- width="76"> . width="71"> ---- width="66"> ---- width="76"> .

R-squared width="71"> ---- width="62"> ---- width="76"> 0.581 width="66"> ---- width="71"> ---- width="76"> 0.581 width="71"> ---- width="66"> ---- width="76"> 0.581

RMSE width="71"> ---- width="62"> ---- width="76"> 2.553 width="66"> ---- width="71"> ---- width="76"> 2.553 width="71"> ---- width="66"> ---- width="76"> 2.553

References

• Achterbosch, T. J., Engler, A., Rau, M. L., & Toledo, T. R. (2009). Measure the measure: the impact of differences in pesticide MRLs on Chilean fruit exports to the EU.
• Ahmad, M. (2007). Impact of Origin Rules for Textiles and Clothing on Developing Countries. ICTSD Programme on Competitiveness and Sustainable Development, International Centre for Trade and Sustainable Development (ICTSD). Geneva. Switzerland
• Bailey, A. E. (1948). Cottonseed and cottonseed products. Their chemistry and chemical technology. Cottonseed and cottonseed products. Their chemistry and chemical technology.
• Bambawale, O. M., Singh, A., Sharma, O. P., Bhosle, B. B., Lavekar, R. C., Dhandapani, A., ... & Pawar, V. M. (2004). under Integrated Pest Management in farmers' participatory field trial in Nanded district, Central India. Current Science, 86(12), 1628
• Bergstrand, J. H. (1990). The Heckscher-Ohlin-Samuelson Model, the Linder Hypothesis, and the Determinants of Bilateral Intra-Industry Trade. Economic Journal 100(4), 12.
• Haq, & Meilke. (2009, 2010). The role of income and non-homothetic preferences in trading differentiated food and beverages: The case of Canada, the United States, and selected EU Countries. Canadian Journal of Agricultural Economics, 57, 169-189.
• Haq, Meilke, K., & Cranfield, J. (2013). Selection bias in a gravity model of agrifood trade. European Review of Agricultural Economics, 40(2), 331-360.
• Jacks, D. S., Meissner, C. M., & Novy, D. (2008). Trade Costs, 1870-2000. The American Economic Review, 98(2), 529-534.
• Kanoi, F. (2004). Agri-Inputs Marketing Research. Cotton Crop Track, A Report on Seeds. Chennai, 2005.
• Krugman, P. (1979). Increasing Returns, Monopolistic Competition, and International Trade. Journal of international Economics, 9(4), 469-479.
• Rau, M. L., Shutes, K., & Schlueter, S. (2010). Index of heterogeneity in requirements in international agri-food trade (No. 10/01). Centre de Coopeération Internationale en Recherche Agronomique pour le Developpement.
• Stalker, H. T., Dhesi, J. S., & Kochert, G. (1995). Genetic diversity within the species Arachis duranensis Krapov. &W. C. Gregory, a possible progenitor of cultivated peanut. Genome, 38(6), 1201-1212.