The empirical study is aimed at understanding the swine marketing ecosystem in selected North Eastern states from the perspective of traders. A cross-sectional survey of 97 traders from Assam Meghalaya and Nagaland was conducted in major and minor markets of rural and urban areas to understand the swine trading process. The mean traders’ income was Rs. 293074 where expanses under variable cost were higher as compared to fixed cost as investment in technology and storage facilities are insignificant.
The price of swine per kg ranges from Rs. 300 to 380 in three selected states of North East. It is determined based on weight which is not weighed through a machine but considered by experience in the trading market. Live pigs are mostly preferred in Meghalaya trading points which range from Rs. 11,500 to 15,000 whereas piglets are traded with a price range between Rs. 3000 to 4800. Price variation at the macro level depends upon the breed and size of the animal. As shown in column 6 in Table 1 the trading preference was found to be different in the region. Higher diversity in the channel was found to be directly correlated to the quantity traded and average income earned. The average year of experience of the traders in the swine business was 8 years with an average income and the total cost was Rs. 2.9 lakhs and 24,428 per year, respectively. The daily footfall of the buyer in the trading market is approximately 53. The range of yearly income in the surveyed area is between 10 thousand to 9.36 lakhs.
The key role in both forward and backward linkages of traders in the swine business of Assam is represented by the chart as shown below. Fig 1 and 2 indicates traders’ participation in the swine business by market chain actor and by value chain actor with price variation. The traders procure the animal from diverse sources like aggregators or middlemen who collect animals from villages, producers and also imports from nearby places or countries, especially from November to April. The aggregators have local villagers in the selected area to give information about the availability of animals and expected prices. Per information they are paid Rs. 100. The increasing demand for swine meat with a declining trend in production in North East states is creating a gap in the trading market. The vulnerability of traders is increasing as they are sourcing animals from countries and regions. Entry of sick animals from nearby regions always adds to the vulnerability of the swine business.
The charts explain the existing procurement chain with price ranges (without value addition and with value addition) in Assam which can be also partially replicated in other states of the region.
Central role of local traders in supply chain
Traders collect animal in Assam from three prime sources (Importing from other states, Producers, Aggregators and large sized producers) and sells to abattoir, retail market and in wholesale market. The marketing efficiency came out to be 1.5, 1.48 and 2.13 for Assam, Meghalaya and Nagaland, respectively
(Bhattacharjee et al., 2022). Fig 2 shows the distribution path of further actors along with the average price ranges which clearly indicates the role of organised value added actors with higher price ranges.
Both charts (Fig 1 and Fig 2) present the role of local traders in the finely branched supply network. It spreads the accessibility of the animal business both with value addition and without it. The customers directly interact with retailers in the daily market/weekday market, Open market (vendors) - Street corner shops in urban and peri-urban areas, small outlets, Super market like Tanz, Big Bazaar, 7 to 9, Orion, NRL Quick shop and with Govt./Pvt., Meat outlets. The availability of products to the customer along with quality and variety initiates by local traders.
Identifying the key components for traders in the swine market: KMO and Barlett’s test
The KMO test value was 0.67 (Table 2) which is larger than 0.6 and found to be acceptable
(Kaiser, 1974). Barlett’s test has a significant value (.000) which explains that the correlation between variables is large enough to be used as factor analysis
(Barlett, 1951), please check. Testing by the correlation matrix was made to show that variables are not correlated whatsoever with each other but relate with themselves. As the data set fulfils the minimum criteria for conducting Principal Component Analysis the extraction of data was further carried out with a total variance table and scree plot as shown below.
Total variance and scree plot
Total variance (Table 3) and scree plot (Fig 3) extracts four components from seventeen as Component 1 to 4 was found to have Eigen value >1. The Kaiser-Guttman rule states that components based on eigen values greater than 1 should be retained. This is based on the notion that since the sum of the eigen values is p an eigen value larger than 1 represents an above-average component
(Kaiser, 1960). The scree plot examines the plot of eigen value and the clear break should occur between meaningful components and a long run of components with eigen values that ‘trail off’ and are meaningless. The components up to the break point are retained
(Raymond Cattell, 1966).
In the scree plot as in the X-axis, 17 components are plotted along with their respective eigen values. Till the 4
th component, the slope is quite significant and later it becomes almost parallel to the X axis with less than 1 eigen value.
The optimal number of components as shown in Table 3 and by the scree plot (Fig 3) is Four. The solution accounts for 67.46 per cent of variance was explained by the four extracted components with the loosing of 33 per cent in other remaining components.
Rotated component matrix
An orthogonal rotation (varimax) is carried out (Table 4) as it does not allow any factors to correlate with each other. It is done to create few variables with optimal weights from a set of variables (17) to capture the variables influencing the swine trading business. To simplify the loadings and to interpret the variables that are most strongly associated with each factor, the following components (Table 5) are named based on extracted variables.
Key components in trading of swine
Component 1 variables like sales or market outlet, com-petitor price and Pig size is strongly grouped under a head which comes under the market description. Component 2 correlates highly with transport cost and labour cost. These items relate to out-of-pocket costs. Component 3 correlates with the marketing plan and quantity of retail cuts and which relates to drive to sale and value addition. Component 4 correlates to Awareness of services which relates to Awareness. The single variable was found to be strongly associated with the rotated component matrix.
Market description, out of pocket cost, drive to sale and value addition and awareness are the components which are named according to the associated components with higher loadings. They are not related to each other. Varimax rotation does not allow the given factors to correlate and therefore these correlations express themselves as cross-loadings.
Factors influencing the income
A multiple linear regression model was fitted to understand the significant factor on which the income of the traders is dependent. Multiple regression is a useful way to quantify the relationship between more than one predictor and one dependent variable. In the study, annual income of the traders is dependent factor and the predictors are the quantity of retail cut, stock retained, Total Variable Cost, No. of buyers purchased, Total quantity sold (live) and transport expenses. To check the fitness of the linear regression model the required tests are done as given below.
Skewness and kurtosis
Values of Skewness and Kurtosis are between 2 and 4, respectively, which was considered acceptable to prove normal distribution. Hair (2010) and Bryne (2010) argued that data is considered to be normal if skewness is between -2 and +2 and kurtosis is between -7 and +7 (Table 6).
Durbin watson statistic
Durbin Watson statistic (Table 7) was used to test the autocorrelation in residuals of a regression model. It checked the pattern of correlation between consecutive residuals (whether they are dependent on each other). The value was 1.5 which was within the threshold. It indicates there was no significant autocorrelation (residuals are independent) between the variables.
Variance Inflation Factor
Value of variance inflation factor was less than 5 (Table 8) which indicates a moderate correlation between variables for 4 variables but it is 7 for stock retained and market sold. Collinearity tolerance of these two variables with higher VIF is higher than 0.1 so it is assumed that multicollinearity does not exist
(Kim, 2019). The predictor values in all three states were measured on drastically different scales, therefore standardized coefficient was considered in the study.
Outliers and Homoscedasticity
With box plot and scatter plot the outliers and Homoscedasticity are checked with fitted value and residual plot. Both fulfil the required condition to fit a linear regression equation.
ANOVA
From the ANOVA (Table 9) the multiple regression model is set as
Y=F(6,90)=18.163 with p-value between .000 to .010, F=18.163 and df=6
Findings of a dependent factor of traders’ annual income:
Finally, from Table 10, the key factors influencing income are given on the basis of significance level. Moreover, R square (Table 9) is 0.548 which depicts that the model explains 54.8 per cent of variance in traders’ income. Approximately 55 per cent of the change in income is significantly dependent on the quantity of retail cut, stock retained and total variable cost. Another factor with relatively less significance is influenced by the number of buyers, quantity sold in the market and transport cost.
Total quantity sold (live) in the market, total variable cost and quantity of retail cuts were significant variables influencing the income of the traders. Here the beta coefficient is negative for the live pig sold in the market and stock retained which indicates that annual income decreases with an increase in stock held and also with dealing in the live pig business. Every one-unit increase in the selling of live pigs in the market and stock retention will reduce income by 0.3 and 0.4 units. So, there is a reverse direction on annual income with live pig selling and stock retention. Qualitative data analysis showed that there is no proper weighing machine or tool to estimate the price of live pigs in the terminal market or daily market. Estimation is made by the raw method- looking at the animal and assuming the weight of the animal. Live pigs are mostly sold in distress or in the piglet stage, which does not give a positive outcome to traders’ income relative to retail cut. In addition, stocks retained, transport expanses (p=.008) and number of buyers purchased (p=0.010) also influence income though stock retained negatively influences income.
From Table 11 and Table 9, it is understood that Meghalaya trades only on live pigs in the terminal market due to the lack of a slaughtering facility and for the lack of proper trading and marketing place. The study site was selected purposefully as pig density, per capita pork consumption and pig trades are highest in Nagaland
(Singh et al., 2021). Nagaland deals both in retail cut and live pig but local traders did not get the opportunity to collect the stock from the village they had to get it from neighbouring states. Assam on the other hand applied the highest mix of trading pathways so the average income of traders goes higher.
The absence of a scientific weighing machine or minimal quality control measures is keeping the entire business highly vulnerable. Literature at various levels has proved that in North East region the pig is not only preferred for rearing but also it is highly consumed. But for the limited importance given to efficient trading methods the swine business is gradually losing its market and it is further getting dependent on animals coming from other states and countries. It creates a dual negative effect. This increases the risk of spread of sickness and makes it costlier too among the consumers.
