Wednesday, April 17, 2013

Elementary misuse of spreadsheet data leaves millions unemployed | Bill Mitchell – billy blog

The blogosphere this morning is all over two researchers, Rogoff and Reinhart, who committed a spreadsheet error (read this as:  messed up their excel file of data) and published based on the file.  This then led to policy recommendations that are based on faulty findings.  R&R said that the tipping point for a country in terms of debt was a 90% ratio of debt to GDP according to their spreadsheet of historic information.  If a debt load grew more than that, the country imploded.  When all the data is included in the spreadsheet, the negative ratio becomes positive.

Earlier this year, the IMF researchers admitted that their estimates of low Keynesian multipliers (1% or less) for fiscal stimulus (changes in G essentially) were wrong.  Multipliers were 1.5% or more. 

Elementary misuse of spreadsheet data leaves millions unemployed | Bill Mitchell – billy blog

1 comment:

  1. Though the article points out the forecasting problems through errors in coding in Excel, the author also points out systematic problems. The author questions the initial economic theory on which the data sets are built upon. "The IMF forecasts are always systematically wrong because they use flawed macroeconomic approaches, which allow ideology to triumph over understanding." I do agree with his warning to not let theory override basic observation, but I do not think marco theory should be ruled out. After the crisis there has been a shift in marco economic thought. Ideally, I would like to a consensus among the economic community. How can we build predictions using data if people can not agree on the initial models to use?

    ReplyDelete