Nothing is more important than data for government statistical agencies. We study statistical classification standards, statistical investigation methods, statistical data collection methods, statistical data processing methods and statistical data evaluation techniques, all in order to obtain true, accurate, complete, timely, representative, scientific, economical and applicable statistical data.
The arrival of the era of big data brings great development opportunities to government statistics, but also brings severe challenges.
First, the application of big data in government statistics.
The National Bureau of Statistics attaches great importance to the application of big data in government statistics. At present, strategic cooperation agreements have been signed with 17 big data companies. Of course, the application of big data in China government statistics is still in its infancy, mainly in two aspects: First, big data has become a part of government statistics sources; Second, big data has become part of the evaluation basis for the quality of government statistical data.
(a) Big data has become a part of government statistical sources.
At present, big data has become a part of the statistical data of China government. The following are several representative aspects:
1. Use data from major online retail trading platforms to measure online retail sales.
In order to master the transaction scale and structure of online retail trading platforms and comprehensively measure the online retail data, the National Bureau of Statistics has conducted a monthly survey of online retail trading platforms since June 5438+ 10 this year, covering 42 key online retail trading platforms such as COFCO Mall, Amazon, Dangdang, Taobao, Tmall Mall, Brewmaster, Meituan, COFCO Womai, Gome Online and Public Comment. According to the data of the above 42 key online retail trading platforms, from June to August this year, the national online retail sales reached 2,240.09 billion yuan, a year-on-year increase of 36.5%. Among them, the online retail sales of physical goods was 65.438+0.86534 billion yuan, an increase of 35.6%, accounting for 83.3% of all online retail sales; The online retail sales of non-physical goods was 374.75 billion yuan, up by 465.438+0. 1%, accounting for 65.438+06.7% of all online retail sales. This has an important reference function for macro-management departments and the public to understand the online retail situation.
2. Calculate the price index of new houses in 70 large and medium-sized cities in China by using the data of housing transaction online signing.
The online signing data of house transaction refers to the relevant information that the real estate development enterprises put on record in the housing management department and published on the real estate information network after the buyers and sellers signed the house purchase contract, including detailed information such as address, floor, price, area and amount, which basically covers all new house transactions in that month. From 20 1 1 to 1, the National Bureau of Statistics began to calculate the price index of new houses in 70 large and medium-sized cities across the country by using online housing transaction data. This has played an important role in improving the data quality of new residential price index in 70 large and medium-sized cities.
3. Use the price information provided by Zhuochuang Information Company to monitor the market price of important means of production in the circulation field.
The National Bureau of Statistics cooperates with Zhuochuang Information Company to monitor the market prices of important means of production in the circulation field by using the price information provided by enterprises. From 20 14 to 10, the market price test results of 50 important means of production in 9 categories in the circulation field were released to the public at the same time. The industry covers ferrous metals, nonferrous metals, chemical products, coal, oil and gas, non-metallic building materials, agricultural products, agricultural means of production, forest products and other fields. Regional monitoring covers 24 provinces, autonomous regions and municipalities including Beijing, Tianjin, Hebei, Shanxi, Inner Mongolia, Liaoning, Jilin and Shanghai. This has played an important role in understanding the market price information of important means of production in the circulation field for macro-management departments and the public.
(b) Big data has become a part of the basis for evaluating the quality of government statistics.
In addition to using big data as a part of the source of government statistics, the National Bureau of Statistics also attaches great importance to using big data to evaluate the quality of government statistics. The following are two representative aspects at present: First, the data quality of the total retail sales of social consumer goods is evaluated by using the consumption data of China UnionPay interbank bank cards; The second is to evaluate the data quality of fixed assets investment by using the Internet of Things data of large machinery and equipment enterprises.
Second, the opportunities and challenges brought by big data to government statistics
For government statistics, big data brings both great development opportunities and severe challenges.
(a) Big data brings significant development opportunities for government statistics.
First of all, big data will continuously improve the macro-management of government statistical services and the public's ability. With the continuous development and improvement of big data and the continuous improvement of the ability of government statistical agencies to develop and apply big data, the types of government statistical products will be continuously enriched, the quality and timeliness of government statistical data will be continuously improved, and the macro management of government statistical services and the public's ability will be continuously improved.
Second, big data will revolutionize government statistics. With the development and perfection of big data and the gradual maturity of the development and application of big data technology by government statistical agencies, government statistics will undergo revolutionary changes. First, the existing statistical survey method system based on regular census, taking sampling survey as the main body, comprehensively using comprehensive survey, key survey and other methods, and making full use of administrative records and other materials may undergo major changes. For a long time, sampling survey method, that is, taking samples from the population and using the samples to calculate the population; Census and comprehensive survey methods, that is, the method of investigating all units one by one in the whole population, have played an important role in the statistics of our government. These methods will still be widely used by government statistics for a long time to come. However, with the continuous development and improvement of big data, big data in some fields and aspects may replace sampling survey, general survey and comprehensive survey methods and become an important method to obtain statistical data, and this method to obtain statistical data will become more and more important. Second, the data collection method of government statistics may change greatly. For a long time, government statistical agencies have mainly collected raw data through enterprise declaration, family bookkeeping and investigator visits. With the continuous development and improvement of big data, some original data will be obtained by mining big data, and this new data collection method will become more and more important. Third, the data processing method of government statistics may undergo major changes. With the continuous development and improvement of big data, the current data processing mode of directly reviewing, summarizing, processing and amplifying sample survey data may undergo major changes.
(b) Big data poses a serious challenge to government statistics.
First, big data challenges the statistical ability of the government. It takes a long time from the emergence of big data itself to its development and perfection, from the initial application of government statistics to the mature application of big data. In this process, on the one hand, the traditional methods of statistical investigation, data collection and data processing in government statistics will continue to operate, otherwise it will not meet the needs of macro-management and the public. On the other hand, the government statistical system must invest a lot of manpower and material resources to mine, process and apply big data, otherwise it will not be able to meet the needs of macro management and the public in the era of big data. This dual-track operation mode will be a great challenge to the government's statistical ability.
Second, big data challenges the traditional concept of government statistics. There is a conventional concept in traditional government statistics: sampling survey method can reduce survey cost and improve efficiency and data quality. Because the sampling survey only investigates some samples in the population, rather than every unit in the population, the number of investigation units is obviously reduced, which can reduce costs, save time and improve efficiency. At the same time, due to the small number of investigation units, government statistical agencies have the ability to provide solid training and guidance to grassroots statistical investigators, and have the energy to conduct strict inspection and audit of statistical survey data, thus improving the quality of statistical survey data. With the continuous development and improvement of big data, more and more government statistical agencies will indirectly obtain statistical data through big data enterprises, without directly investigating specific units as a whole or training investigators. The advantages of low survey cost and improving the quality of statistical survey data will no longer exist.