Optimization of distribution network in the GST regime: An Indian perspective

Abstract

With an aim of simplifying the indirect tax regime and improving the business ecosystem, the Indian government is trying to implement the Goods and Services tax (GST) which subsumes most of the indirect taxes currently in effect. This tax is slated to have major impact on the distribution network of Indian firms. Currently, most of the Indian firms’ distribution networks are designed for reducing the tax burden on the products rather than improving the logistic efficiency of the distribution network as a whole. The multiplicity of taxes, complex tax jurisdiction and legislation have all proved a hindrance to firms trying to improve logistic efficiency. GST with its single tax and clear taxation jurisdiction and policy aims to clear up the confusing tax landscape in India. Also, the simplified tax policy gives firms to model tax liability accurately which wasn’t possible because of the lack of clear tax policy which made tax calculation and distribution optimization for firms with national level distribution. Taking cue from the simplified tax structure, in this paper, we identify the effects of GST on the cost structure of the firm and model the distribution costs of firms dealing in multiple products on a national scale. We modeled the overall distribution tax structure using the five major components of distribution networks in India, i.e.: safety stock inventory costs, transit inventory costs, transportation costs, fixed costs of setting up and operating distribution centers and the tax costs resulting out of distribution choices. We then developed an optimizing problem to minimize the total distribution costs by taking the number of distribution centers and the servicing sequence of the demand points as dependent variables. We developed a non-linear mixed integer – programming model to optimize the distribution costs and distribution center locations with predefined constraints. This model was then examined our analytical model for possible trends by variable variation method and the macro trends that would be visible in firms’ distribution network are obtained. Also, we performed a significance and impact check to identify the impact that each variable of the distribution cost model will have on the overall distribution network. These results form the core part of the objectives of this project. Also, in addition, we developed a spreadsheet model and a mathematical code (using MATLAB) which can be used by firms with specific parameter set to obtain the optimal distribution network for their specific use cases.