TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Reinhart-Rogoff Response to Critique

55 点作者 fretlessjazz大约 12 年前

5 条评论

doktrin大约 12 年前
&#62; <i>...these strong similarities are not what these authors choose to emphasize.</i><p>This statement feels misleading, in that the similarities do not appear to be as significant as they are implying. The table they provide very clearly shows a significant difference between their 2010 findings and the HAP paper (-0.1 vs. 2.2 mean, a gap of over 2 percentage points).<p>&#62; <i>These results are, in fact, of a similar order of magnitude to the detailed country by country results we present in table 1 of the AER paper</i><p>Taken literally, this makes no sense. No one has ever claimed the HAP results were of a differing <i>order of magnitude</i> to their own.<p>Given that this is a rebuttal to a research paper where numerical values are in dispute, I would have thought they would use the term ("order of magnitude") precisely, and not colloquially.<p>&#62; <i>It is utterly misleading to speak of a 1% growth differential that lasts 10-25 years as small.</i><p>Perhaps, but if a 1% growth differential is significant (HAP), then a 2.9% differential is <i>massive</i> (2010 RR), compounding the scale of their error.<p>After all, their original 2010 work implied a mean 2.9% drop in the growth rate once debt climbed above 90%. That paints a significantly different picture than a decrease of 1 percentage point, resulting net positive growth rate (2.2%) as opposed to negative (-0.1%).<p>While this is speculative on my part, I feel it's safe to say that had they originally reported these (allegedly) "very similar" numbers ( &#62; 2% vs -0.1 %), their findings would not have gotten the same wide circulation in certain political circles that it did. Given that context, the whole "<i>...but the numbers are kinda close...</i>" argument falls a little flat IMHO.
评论 #5561747 未加载
anigbrowl大约 12 年前
(copied from my comment on the article)<p>The original paper concluded that debt overhang above 90% of GDP would result on growth of -0.1%. This was based on three serious flaws – 2 of methodology, one of spreadsheet programming. The corrected figure is 2.2%. To pretend that this is not a big deal is disingenuous considering how widely the 2010 paper was cited. It’s true that the 2012 paper is more accurate and also that higher debt leads to lower growth, but to gloss over the severe flaws of the earlier paper is misleading. I would respect Reinhart and Rogoff a lot more for simply acknowledging the errors and repudiating their previous conclusion; instead they essentially argue their errors ought to have been caught earlier.
评论 #5562243 未加载
GabrielF00大约 12 年前
This is an unsatisfying response. It doesn't answer the question of why Reinhart and Rogoff chose to exclude certain high-debt years for certain countries or why they chose the particular weighting mechanism that the critics found problematic.
评论 #5561779 未加载
评论 #5562744 未加载
评论 #5561886 未加载
gruseom大约 12 年前
Dean Baker's critique of this response is typically incisive:<p><i>[Herndon, Ash, and Pollin] found growth was slower in periods with debt levels above 90 percent of GDP than below, but the gap was relatively small and nowhere close to statistically significant. Furthermore, they found a much bigger gap in growth rates around debt-to-GDP ratios of 30 percent. If we think that [Reinhart and Rogoff's] methodology is telling us something important about the world then the take-away should be that we want to keep debt-to-GDP ratios below 30 percent.</i><p><a href="http://www.cepr.net/index.php/blogs/beat-the-press/quick-thoughts-on-reinhart-and-rogoffs-response" rel="nofollow">http://www.cepr.net/index.php/blogs/beat-the-press/quick-tho...</a>
评论 #5565117 未加载
jacquesm大约 12 年前
A Wolfe Simon worthy response.