Tag Archives: Statistics

Kevin Lang argues in NBER wp 31666: When economists analyze a well-conducted RCT or natural experiment and find a statistically significant effect, they conclude the null of no effect is unlikely to be true. But how frequently is this conclusion warranted? The answer depends on the proportion of tested nulls that are true and the power of the tests. I model the distribution of t-statistics in leading economics journals. Using my preferred model, 65% of narrowly rejected null hypotheses...

A reminder not to be overly impressed when presented with statistically significant coefficients, from FiveThirtyEight.com.

China’s export growth disappointed in July, only we don’t really know by how much. According to that country’s Customs Bureau, exports last month were 7.2% above (in US$ terms) exports in July 2016. That’s down from 11.3% growth in June, which as usual had been taken in the mainstream as evidence of “strong” or “robust” global demand. According to China’s National Bureau of Statistics, however, exports in June rose by...

Chinese trade statistics were for May 2017 better than expected by economists, but on the export side questions remain as to their accuracy. Earlier this year discrepancies between estimates first published by the General Administration of Customs (GAC), those you find reported in the media, and what is captured by the National Bureau of Statistics (NBS), backed up by data from the Ministry of Commerce, became...

Chinese economic data for April 2017 has been uniformly disappointing. External trade numbers resembled too much commodity prices, leaving an emphasis on them rather than actual economic forces. The latest figures for the Big 3, Industrial Production, Retail Sales, and Fixed Asset Investment, unfortunately also remained true to the pattern. Industrial Production had seemingly accelerated in March, rising to a 7.6%...

China’s trade statistics for April 2017 uniformly disappointed. They only did so, however, because expectations are being calibrated as if the current economy is actually different. It is instead merely swinging between bouts of contraction and low-grade growth, but so low-grade it really doesn’t qualify as growth. Positive numbers do get the mind racing, but since the end of 2011 there is almost a speed limit on how...

As with all statistics, there are discrepancies that from time to time may obscure the meaning or validity of the particular estimate in question. For the vast majority of the time, any such uncertainties amount to very little. Overall, harmony among the major accounts reduces the signal noise from any one featuring a significant inconsistency. There are, of course, various economic areas where estimates are going to...

In separate blog posts, Russ Roberts and John Cochrane have called for humility on the part of economists. Asking “What do economists know?,” Roberts and Cochrane point out—correctly—that economics is not as strong on quantification as some economists and many pseudo economists pretend, and as is often expected from economists. Economics is not the same as applied statistics although the latter can help clarify, at least to some extent, the empirical relevance of economic theories....

The National Bureau of Statistics (NBS) of China does not publish any of the big three data series (Industrial Production, Retail Sales, Fixed Asset Investment) for the month of January. It combines January data with February data because of the large distortions caused by Lunar New Year holidays. Unlike Western holidays that are but a single day, the Golden Week is a week, and therefore when the calendar points do not...

The Economist recounts the reasons why the quality of GDP measurement is lacking and why GDP measures cannot answer all the questions they are used to address.