iotables: an r Package for Reproducible Input-Output Economics Analysis, Economic and Environmental Impact Assessment with Empirical Data

The supply and use tables (SUTs) and input-output tables (in short: IOTs) provide a very detailed, and empirically measured or estimated picture of the economy.  The tables present information on the supply and use of goods and services for industries’ intermediate consumption and categories of final use (final consumption, capital formation and exports). They also provide details on the generation of income for each industry distinguishing the components of gross value added. The SUTs and IOTs provide empirical data for a wide range of economic analyses. 

These tables show the inter-linkages of a national or regional economy both from a supply and demand perspective. Where various sectors of the economy typically purchase goods and services from each other as they work themselves towards the final consumers or towards exporting outside of the economy described with the data. How the economy distributes corporate, personal and government income, how they finance consumption and investments. They system of input-output tables are the most comprehensive, empirically measured data for many types of macroeconomic research, industry organization analysis, or they provide tools for various economic and environmental impact analysis.

The `eurostat::get_eurostat`  function retrieves the relevant supply, use, symmetric-input output or margin tables in a tidy, long-form. These tidy, long datasets cannot be meaningfully spread into a tidy, wide format without understanding and properly ordering the rows and columns of those datasets. The input-output system is a matrix algebraic system. The system of the input-output tables must be spread into at least four, interconnecting and compatible matrixes. Any further data for analysis (such as data on employment, or material flows like greenhouse gases) must be added to a system of matrix equations in the form of conforming vectors or matrixes.  Properly formed coefficient matrixes must be calculated from parts of the input-output table, and they must be translated into the Leontief matrix and its inverse. The *iotables* R package adds the functionality to *eurostat* to properly process the long-form data into many tables---in some cases, the bulk downloader returns more than 800 SUTs in one single long-form dataset.