Editor: Peng Wenhua
Source: Big Data Architect
Hello Peng friends, I am Lao Peng. Two days ago, it was rumored everywhere on the Internet that China's unemployment rate has reached an astonishing 18%! ! ! I'm all blinded! Is this data so amazing?
I was so scared that I checked the source of the data all night. When I said that it was released by the "state", it seemed that it was muddled. There must be a department for data release, and it would not be issued by the "state". This is the meaning of the existence of the responsible department of the data.
The other said that it was released by the National Bureau of Statistics, and it was the data for April. This has a nose and an eye. I took a closer look.
Data Sources
Generally, such data are released by the National Bureau of Statistics. Let's Baidu the National Bureau of Statistics, the first link is:
After opening it looks like this:
See the "Data Query" above? Click to enter. Well, there will be a warning, let’s ignore it and just continue, so there’s no need to complain about it~~~
After entering it is like this:
Yes, clicking "Monthly Data" opens the result of the CPI (Price Index) indicator by default. Sure enough, the data for April 2022 has been updated.
Here is an example of a very nice dataset product. Very standardized and of great reference value. Take CPI as an example:
The leftmost one is the indicator system, which is in a tree-like directory structure. According to the indicator classification standard, tens of thousands of indicators are classified and collected;
On the upper right is the toolbar. Anyone who has made a report knows what is going on, so I will skip this one.
In the middle right is the report, which is the specific data. Those who have done this in excel can understand what it means.
The bottom right is the annotation of the data, which generally explains the cause and effect of the indicators, the relationship between them, and the statistical caliber.
unemployment rate
Our topic today is the unemployment rate. Go directly to the search bar above and enter "unemployment rate" to check:
There are two statistical methods for the unemployment rate: one is the registered unemployment rate, and the other is the surveyed unemployment rate.
Most of the foreign countries use the unemployment rate survey, which means to go to the street to find someone randomly and ask: "Are you unemployed?" This method is more active and closer to the actual situation.
Our country has been using the method of registration and statistics before. This method is relatively passive. Only those who go to social security to register as unemployed will be counted. Therefore, the unemployment rate is generally lower than the surveyed unemployment rate.
In the past two years, we have also started to use surveys to calculate the unemployment rate. Judging from the results, it is indeed higher than the registered unemployment rate. The national urban survey unemployment rate in April was 6.1%, but it is not as high as 18%. !
Lao Peng opened it one by one, and the data is here:
Well, the name of the indicator is "National Urban Survey Unemployment Rate of 16-24 Population (%), as high as 18.2%. The rumors are right, but not all right.
From the above table, we can see that the post-00s Xiao Pengyou is super brave. The 
online rumors about post-00s rectifying the workplace are not groundless~~
As a post-80s me, why do I feel hurt when I am offended 
? Who can I tell?
Has the post-00s been so brave? Or is it braver this year? Lao Peng, I stretched the data to 36 months. Let's ignore the inconsistency of the labels on the left, there are some minor problems with automatic products.
The green line has always been high. Before the epidemic, the national surveyed unemployment rate (orange) was only 5%, and the national urban surveyed unemployment rate (green) of the population aged 16-24 has already rushed to 10.5%!
By the end of 2021, the unemployment rate of post-00s has soared to 15%, to 16% in March this year, and to 18.2% in April. Brave, really brave!
For a moment, Lao Peng was also a little excited.
understand the data correctly
I know that Peng You will come forward to question the accuracy of the data, and even bring up various cases of data falsification.
There will also be people who deliberately abandon the overall situation and take out some data and over-interpret it. Before Lao Peng heard that "the national unemployment rate has reached 18%" is typical.
Others question whether the survey data and calculation process are scientific and rigorous.
However, Lao Peng wants to say that these are not important.
There is no need to wash the ground, and Lao Peng can't come up with a formula now, let alone ask about the employment situation of the 1.4 billion Peng friends across the country. But we can do at least one thing:
Seriously, scientifically, rationally, and accurately treat the data, don’t over-interpret, don’t take it out of context, don’t make random associations, big things will happen
This reminds me that there is a special chapter on data ethics in the "Chief Data Officer Management Manual", and Chapter 2 in "DAMA Data Management Knowledge System Guide DMBOK 2.0" is called "Data Processing Ethics". It all talks about the correct attitude towards data.
But it's full of misquotes, wanton misuse, and even deliberate misinterpretation of data. Alas, the road is long and difficult...
More exciting:
Lao Peng's new book: "The Chief Data Officer Management Manual" is now accepting reservations~~~
What is the difference between data classification and labeling?
Hot thread: Data warehouse is really boring...
We can't help you with the matter of Xi'an One Code, so let's not make trouble?
Compared with DCMM, why does DMBOK have no data standard?
Typesetting | Lao Peng
Reviewer | Lao Peng Editor-in-Chief | Lao Peng