Error Discussion

Hello Peter, We are not aware of any documents that exactly match your criteria. Statistical accuracy implies having a number of observations and measuring how much the observations vary from one another and how well a model is able to explain the variation. I-O is not a probabilistic model but rather a system of accounting from which we make the assumption that the current economic structure holds for any marginal change. Impacts are based on the picture of the economy and the relationships as part of the picture; thus, the predicted change is essentially a product of accounting. The accuracy of IMPLAN's data and analyses is closely tied to the accuracy of our source data, the vast majority of which comes from government sources. This webpage contains links to articles detailing our data sources and procedures: http://implan.com/index.php?option=com_content&view=article&id=414&Itemid=1878 Likewise Indirect and Induced jobs are very hard to verify so it is hard to judge the accuracy of IMPLAN’s estimates of those. However, here is one study that attempts to do just that: http://ageconsearch.umn.edu/bitstream/14669/1/21020231.pdf We have also attached a document that compares our data estimates to three major raw data sources (Census CBP, BLS CEW, and BEA REA). [attachment=654]h71312eb.docx[/attachment] We hope this helps – please let us know if we can provide additional information.