Economic sciences/2.International activity

Natalia M. Lashkevich

Master of Economics, PhD-candidate of the University “G.d’Annunzio” Chieti-Pescara, Italy e-mail: lashkevich.n@gmail.com, priorpost@mail.ru.

Assessment of wheat import supplies and food risks

at the EU single market

Summary

The article presents analyses of cause relations between the EU’ wheat imports and identified food risks to check a hypothesis and to assign the more “harmful supply” by month. The following methods were used: statistical analysis, regression, semi-quantitative assessment.

Keywords: wheat, import, cultivate period, regression, semi-quantitative assessment.

Introduction

The analysis of world wheat import structure showed that the EU-27 takes the first place in wheat imports during last five years. [1] Taking into consideration that the import of harmful and danger grain will lead to the production foodstuffs with food hazards, it is necessary to analysis cause relations between wheat imports and food risks considered cultivate periods.

Statistical analysis

Spring and winter wheat have different cultivate periods.[2], [3], [4]

Table 1 Cultivate periods for spring and winter wheat

Cultivate stage	Month
	Spring wheat	Winter wheat
Sowing	May (the beginning).	The end of August – the beginning of September.
Growing	June – July (generally). The vegetative phase is 75-110 days.	September – June (generally). The vegetative phase is 240-320 days.
Harvesting	The beginning of July – the end of August or the beginning of September.	The beginning of July.
Transport Handling	After harvesting.	After harvesting.
Storage	After handling – generally since September till April/May next year.	After handling – generally since July till August next year.

Source: [2], [3], [4]

So, for the analysis tendencies at the Wheat Markets and directions of imports flows of grains it is necessary to compare Cultivate Calendar for Wheat, food risks and wheat imports into the EU market.

Table 2 Basic data

Month	Cultivate period		Food risks		Wheat imports
	Winter wheat	Spring wheat	QNT	average	millions tons	average
January	Growing		4	4 (incl. Dec)	2,5	2,36 (incl. Dec)
February	Growing		3		2,0
March	Growing		5	3	2,1	2,13
April	Growing		2		1,8
May	Growing	Sowing	2		2,5
June	Growing	Growing	0	0,33	2,0	2,16
	Harvesting
July	Harvesting	Growing	0		2,3
		Harvesting
August	Sowing	Harvesting	1		2,2
September	Sowing	Harvesting	3	2,33	3,8	3,23
	Growing
October	Growing		4		3,2
November	Growing		0		2,7
December	Growing		5		2,6

Source: [5], [6]

Figure 1. Dynamics of food risks and wheat imports, 2001-2011years

Figure 1 demonstrates fluctuations of food risks and wheat imports. The visual analysis showed that during trade period spring-summer the volumes of imports are at medium level. In autumn and winter imports were increased (especially in September – November). The analysis of food risks showed the fluctuation of food risks has the following tendencies: absence in June, July, November; the increase since September; constantly notifications in a period “autumn – spring”; the most quantity of notifications was received in winter.

Regression analysis

Taking into consideration that the harvesting of spring and winter wheat begins in June (as usually), so linear regression will built for calendar period: summer-autumn-winter-spring. The equation of linear regression is

y_(x) = 1,727+0,046x, where (1)

x_j - a single predictor variable – wheat imports by month; y - a response variable – food risks; α – a coefficient – 0,491; β – a coefficient, a slope of the regression line – 0,778.

The coefficients of the linear regression demonstrate that the increase of wheat imports in a month on 0,778MTs will lead to the growth of food risks on 0,491.

The correlation coefficient r_xy is 0,233. So, the linear regression (1) has a positive and weak linear relationship (correlation) between variables. [7] The coefficient of determination, denoted R₂, is 0,0545. The coefficient R₂ converges to the zero and the linear regression (1) cannot be used for the forecast. The value R₂ = 0,0545 may be interpreted as follows: approximately 5,45% of the variation in the response variable can be explained by the explanatory variable. The remaining 94,55% can be explained by unknown, lurking variables or inherent variability. [8] To assess the quality of the linear regression for the forecast, the approximation error is calculated. The recommend level is 8-15%. The calculated approximation error Ā is 31.16%, so the linear regression (1) cannot be used for the forecast.

For the testing null hypothesis the following coefficients were calculated: F-test and student’s t-test with probability α (0,05); k₁ and k₂ degrees of freedom (1; 10 respectively). Taking into consideration that F-calculated < F-statistic (0,57<4,96), the linear regression (1) is significantly at level α (0,05) and t-calculated is below then t-statistic (0,75<2,23) the null hypothesis is rejected in favor of alternative hypothesis. [9], [10]

Semi-quantitative assessment

During the analysis period – 2001-2011 years the following notifications about the presence hazards in the wheat supplies were received: Mycotoxins, Insects, Pesticides residues, Heavy metals, Poor or insufficient controls. Taking into consideration, that in a month some notification were received, so it is necessary to implement scaling assessment by weight number.

Table 3 Scalling system for hazards

Food risks	IARC group	Weight number
Mycotoxins	1	0,3
Heavy metals	1	0,3
Pesticide residues	2B	0,2
Biocontaminants	3	0,15
Poor or insufficient controls	3	0,15
Total		1,0

Table 3 shows that such carcinogenic food risks as” Mycotoxins”, “Heavy metals” and “Pesticide residues” received the following IARC groups as 1, 1, 2B by the principle: a carcinogenic food risk gets number of the more harmful hazard into its categorical group. Among mentioned food risks groups the following hazards were presented: Aflatoxins (IARC group – 1); Dichlorvos (IARC group – 2B); Nickel ore (IARC group – 1). For the identification “harmful” group for non-carcinogenic food risks as “Biocontminants” and “Poor or insufficient controls” used the principle: a carcinogenic food risk gets number of the less harmful hazard into its categorical group. So, these food risks received a IARC group 3, because such hazards as “Insects”, “Labelling” and “Preservation” are non-carcinogenic. The calculation results is presented in table 4.

Table 4 Calculation of weight coefficients

Month	Food risks	Mycotoxins	Insects	Pesticides residues	Heavy metals	Controls	Weight coefficient
	Category	0,3	0,15	0,2	0,3	0,15	Formula
January	Mycotoxins, Poor or insufficient controls	3				1	3*0,3+1*0.15=1,05
February	Mycotoxins	3					3*0,3=0,9
March	Mycotoxins	5					5*0,3=1,5
April	Mycotoxins, Insects	1	1				1*0,3+1*0,15=0,45
May	Mycotoxins, Poor or insufficient controls	1				1	1*0,3+1*0,15=0,45
June							0
July							0
August	Mycotoxins	1					1*0,3=0,3
September	Mycotoxins	3					3*0,3=0,9
October	Mycotoxins, Insects, Pesticides residues	1	2	1			1*0,3+2*0,15+1*0,2 =0,8
November							0
December	Mycotoxins, Pesticides residues, Heavy metals	3		1	1		3*0,3+1*0,2+1*0,3 =1,4

In according with table 4 the interval of weight coefficients is between 0-1,5 thus for the semi-quantitative assessment of food risks and wheat imports will apply the 6-score system presented in table 5. [11]

Table 5 Scoring systems

Score	Category	Import range, MTs	Weight coefficient
6	Very high	3-4	1,3-1,5
5	High	2-2,9	1,0-1,2
4	Medium	1,1-1,9	0,7-0,9
3	Low	0,6-1,0	0,4-0,6
2	Very low	0,1-0,5	0,1-0,3
1	None	0	0

Summarizing scoring systems for every month is presented in scheme 1.

Food risks	Very high					March, December
	High					January
	Medium					February	September, October
	Low				April	May
	Very low					August
	None					June, July, November
		None	Very low	Low	Medium	High	Very high
		Wheat Imports

Scheme 1 Severity level of import period

Scheme 1 allowed to rank months by severity level: Very high - March, December; High -January, February, September, October; Medium -April, May, August; Low - June, July, November.

Hence, the strengthening border controls should be increased in period autumn-winter-spring because during the storage the quantity of food risks is increased (as usually) and admittance of harmful seeds will decrease the future harvests and lead to the expansion of transferred food risks.

Conclusions

Spring and winter wheat has various cultivated periods. The vegetative phase of spring wheat is 75-110 days and of winter is 240-320 days. To analysis the possibility of relationship between food risks and wheat imports by month of supplies the following analysis were applied: statistical and regression analyses, semi-quantitative assessment.

The built linear regression (1) has a positive and weak linear relationship (correlation) between variables: the increase of wheat imports in a month on 0,778 MTs will lead to the growth of food risks on 0,491. However, this regression model cannot be used for the modeling and forecasting because the coefficient of determination 0,0545, thus the remaining 94,55% can be explained by unknown, lurking variables or inherent variability.

Semi-quantitative assessment allowed to rank supplies by “more harmful” month. So, it is necessary to strengthen the border control activity in period “autumn-winter-the beginning of spring”.

References

[1] Grain: World Markets and Trade. Trade Policy Review US DA, FG 10-10. WT/TPR/S/214/Rev.1 8 June 2009 (09-2701) // WORLD TRADE ORGANIZATION

[2] The wheat growth guide, Spring 2008, second edition, 3-6 HGCA [e-source] http://www.goldsino.ru/field-of-activity

[3] Яровая пшеница. [e-source] http://www.fadr.msu.ru/rin/crops/summertriticum1.htm

[4] Озимая пшеница. [e-source] [http://www.goldsino.ru/field-of-activity]

[5] Notifications list [e-source] https://webgate.ec.europa.eu/rasffwindow/portal/

[6] International trade. Eurostat. [e-source] http://epp.eurostat.ec.europa.eu/portal/page/portal/international_trade/data/database

[7] Aitken Alexander Craig ‘Statistical Mathematics’ - 8th Edition. // Oliver & Boyd, 1957. ISBN 9780050013007

[8] Draper N.R. and Smith H. ‘Applied Regression Analysis’ // Wiley-Interscience, 1998. ISBN 0-471-17082-8

[9] Triola Mario ‘Elementary statistics’ (8 ed.). Boston: Addison-Wesley, 2001. ISBN 0-201-61477-4.

[10] Lomax R. G.. ‘Statistical Concepts’: A Second Course, 2007 ISBN 0-8058-5850-4

[11] Exposure assessment of microbiological hazards in food. Guidelines: Microbiological risk assessment series 7. // WHO/FAO, 2008