Over the years, people have raised their hands to ask me, "is the GDP, the cpi, the unemployment rate that you've published?" it's not like it's trueStatisticsHot spots never add to or subtract from a few numbers. They're behind themDataTechnical work on methods of collection, calibration of surveys and international statistical standards to which these ordinary people are rarely exposed. Today, i've been doing this for yearsDisputesThe biggest questions were broken and covered。
What about statistics on hot spots
In terms of GDP, the 2024 round of the economic census replaced the previous sampled services product with full verification, so it was not the same as the previous number, and someone yelled "watering." it's not water. It's statistical coverage. For example, you can't say that kids suddenly become smart when a class that used to take math and now added english and sports. The core logic is:The base figure has changed. It's not fakeI don't know。
Cpi is also a disaster area. It's like, "i've got five points around me before you say zero." the problem arises from the fact that the eight categories of goods and services covered by the cpi are given a five-year shift in weight, with the rigid spending on food, housing, education and young people's most expensive entertainment, digital and out-of-home meals being less important. You feel "high" and "high" in the basket of statistical officesIt's not the same as sampling objectsIt makes no sense。
Areas covered by statistical hotspots
Employment statistics were the worst in two years. Since 2023, unemployment rates have been investigated in parallel with registered unemployment rates, the former being a sample of households asking if you have been looking for a job in the last month, and the latter being actively registered in the street window. The two figures were two or three percentage points short, and the comment area exploded. But..MethodologyIt's different in itself, one for intent, one for behavior, and one to pull it together, like counting reading rates with "you're going to read a few books this year" and "you're actually going to the library."。
Further, the statistical hotspots are tied to the balance-of-payments balance sheet, the release rhythm of foreign exchange reserves. Foreign exchange reserves are updated on a quarterly basis, with a time lag between bank deposits and the exchange rate of the current month. This was not a match during the accounting period and not a data error. It's easy to fall into an emotional trap when you can't see it。
It is not expected that an article will give all the technical details, but it is hoped that next time you get to the "data fighting" post, you can take 10 seconds to ask, "what's the caliber, how are the samples drawn, what's against which standard?" the number itself does not lieTake the presetInterpretationYeahI don't know。
