"So the Reinhart-Rogoff fiasco needs to be seen in the broader context of austerity mania: the obviously intense desire of policy makers, politicians and pundits across the Western world to turn their backs on the unemployed and instead use the economic crisis as an excuse to slash social programs."<p>Amen. After reading any piece of news, you'd think that the only utility economists have is to (rather poorly) justify decisions that are actually fully ideological in nature, using numbers whose origin is uncertain.
Please, bear in mind that Krugman has a very opinionated view against austerity (he seems to be going the Keynes way these days) and this article is not about Excel. He just wants to show how right he is and how sloppy the austerity guys are.<p>The way to compete against Excel is actually learning to love the tool (it took me 2 years, but I almost can say I love it... sometimes).<p>EDIT
Since I am on HN, I always try to have a serious discussion about Excel and business software. That's why I always try to tie this kind of topic to my general idea: that we need to understand very well the type of work people do in Excel before proposing any solution.<p>But... if we want to talk about austerity, let's talk.<p>What is the real problem with anti-austerity measures today? We could accept a GDP/Debt ratio of 200, 300, 1000, infinity, and let people live their lives, right? Well, hell no. For 1 simple reason: interests. And I am not talking about the obvious problem with interests (that you have to pay them), but the bit that everybody forgets: they are a bet against the future that growth will be high enough to pay them. And that's the problem. We have a low growth rate expectancy (we western countries). And that's why serious economists (meaning: not Krugman) are telling us that increasing our debt exposure is the path to future ruin.<p>Now, you will say, but anti-austerity financed with higher taxes will bring growth! If that were true, we would have had serious growth the last 50 years (a period within which tax ratio on GDP increased pretty much everywere). The truth is that historically public spending has brought much less growth (if any) than private spending. Meaning: for each $ that we take away from privates to the public we are deliberately reducing present and future growth. And why public spending brings less (if any) growth? Well, simply because the market is so complex that we need multiple agents. Giving too many resources to 1 agent (the State) is like betting most of our money on 1 black jack game, does it make sense? No. You need to spread risk and resources, you need to let people make decisions and you need to let people keep most of what they produce.
GDP is in $/yr, while debt is in $. Thus, debt/GDP has units of time. 90% really means 10.8 months.<p>One of the reasons the idea was sticky is that 90% seemed like a significant number, almost all the way to 100%. Where there's nothing so obviously magical about 10 or 12 months.
Question: has Paul Krugman made a practice of releasing his raw data? Can anyone point me to the database of tab-delimited text files behind the econometric papers he's written, or to any systematic attempts to replicate them? If not, should be fun to try to get a dataset out of him. Most published macroeconomics research won't stand up to attempts to replicate it (whether "pro" or "anti" austerity), if the great replication experiment in psychology is any guide:<p><a href="http://psychfiledrawer.org" rel="nofollow">http://psychfiledrawer.org</a><p>Basic kinds of corrections like weighting nations by population will often invalidate macroeconomic correlations.
The excel error:
<i>Coding Error. As Herndon-Ash-Pollin puts it: "A coding error in the RR working spreadsheet entirely excludes five countries, Australia, Austria, Belgium, Canada, and Denmark, from the analysis. [Reinhart-Rogoff] averaged cells in lines 30 to 44 instead of lines 30 to 49.</i><p><a href="http://www.nextnewdeal.net/rortybomb/researchers-finally-replicated-reinhart-rogoff-and-there-are-serious-problems" rel="nofollow">http://www.nextnewdeal.net/rortybomb/researchers-finally-rep...</a>
I thought it was kind of well known: don't use Excel for floating point math, statistical analysis, anything moderately involved that would have to be reproducible and auditable<p><a href="http://www.pages.drexel.edu/~bdm25/excel2007.pdf" rel="nofollow">http://www.pages.drexel.edu/~bdm25/excel2007.pdf</a><p><a href="http://support.microsoft.com/kb/78113" rel="nofollow">http://support.microsoft.com/kb/78113</a><p><a href="http://en.wikipedia.org/wiki/Numeric_precision_in_Microsoft_Excel" rel="nofollow">http://en.wikipedia.org/wiki/Numeric_precision_in_Microsoft_...</a>
I don't really understand how this paper could be claimed to have 'destroyed the economies of the world' when it's in the aftermath of the GFC. 'Not made it better', sure, but bad fiscal practices causing deep financial problems were there before 2010.
Why not the 'Arithmetic Error Depression' in place of 'Excel'?<p>To accurate/concise for Krugman? Must be half a dozen articles about this I've seen now and all have titles suggesting Excel made the error.