China - Consumer Price Index (CPI)

The consumer price index (CPI) measures changes in the prices of goods and services consumed by households. Such changes affect the real purchasing power of consumers’ incomes and their welfare. In practice, the official CPI of China measures the change in the level of prices for a specified basket of goods and services normally purchased by urban and rural residents.

Active:

• Measurements:
  • Fixed-base index relative to 2015 (Index 2015=100)
  • Moving-base index "same month of prior year=100" (Index CPPY=100)
  • Moving-base index (prev. month=100)
  • Year-over-year percent change (% Y/Y)
• Adjustment: Not seasonally adjusted (NSA)
• Native frequency: Monthly
• Start date: As early as 1987m1
• Geo coverage:
  • Country
  • 31 provinces

Predecessors:

• 2010=100, 1986 to 2015

The CPI survey covers around 500 areas throughout the country, including around 200 counties and around 300 cities. The current CPI is a chained Laspeyres price index, which is compiled and disseminated since 2001. Prior to 2001, the index was compiled and disseminated on a month to month and year to year basis using current year weights for most items.

262 item groups have been selected as representative items of consumer purchases from the household surveys. Each representative item consists of around 2-10 varieties, for certain special items have 15-25 varieties each.

Prices for most items are collected at least two times during the month in shops and markets in selected areas. Prices of perishable items – fruits, vegetables, eggs, poultry, meat and fish – are collected six times every month and prices of industrial products are collected two times a month. Where prices are still supervised by the government, such as rent, electricity, public transportation, and drinking water the information on such prices is collected once a month.

Items that are no longer available are replaced by similar items.

Subnational data is available for total CPI previous month=100. Geo coverage includes all 31 provinces in China.

The source writes:

Beginning in January 2016, 2015 was used as a base period in the new round of rotation, and the base periods in the previous three rounds were 2000, 2005 and 2010. Compared with the base period in the last round, the survey directory and specifications of the new base period were closer to international standards, and a number of new products and services were included, which could further reflect changes in consumption and economic structure. At the same time, the weight numbers of CPI composition were adjusted accordingly, according to the latest results of the survey on residents income and expenditure. The survey directory of the new base period was revised following the Classification of Individual Consumption According to Purpose (COICOP) developed by the United Nations and the Classification of Household Consumption Expenditure (2013) enacted by the NBS. According to estimation, the current rotation of the base period had almost no influence on the month-on-month data of CPI.

Moody's Analytics supplements

We back-extend the NSA general index.

We compute a seasonally adjusted version of the general index.

For the percent-change variant, we back-extend to align with the indexed variant.

Data are not revised.

The source writes: "Article 6 forbid revising the data provided by the NBS and prohibit officials from influencing, altering or tampering with the data. The NBS is held solely responsible for the authenticity of the data produced and disseminated by it."

Translation to Data Buffet

The source (NBS) publishes two analytic variants of the moving-base CPI, titled in English as follows. We harvest only the first.

"Same month previous year=100" is a non-cumulative ("discrete") value, which we label with the unit-descriptor metadata (Index CPPY=100, NSA). For example, CPITTYUM.ICHN.

Conversely, "same period previous year=100" is cumulative, and we do not harvest it. If we did, its metadata would be (Index CPPY=100 YTD, NSA).

Interpretation

For monthly consumer prices, we harvest both the "index, same month previous year" and "percent change, year-over-year" analytic variants. They are numerically related in the following way. For example, for the all-items branch:

CPITTYUM.ICHN - 100 = CP%TTYUM.ICHN

Dismal Scientist / Economic View

For the percent-change figures (CP%{TTY,CLTY,FPBY,HOSY}UM.ICHN), the most recent period is reported by the source, but its history is Moody's Analytics Calculated.

For the indexed figures (CPI{TTY,CLTY,FPBY,HOSY}UM.ICHN), the most recent period is Moody's Analytics Calculated, and the history is reported by the source.

Further reading

At the source:

• Rebasing (Jan 2018, Chinese)

At IMF (SDDS):

• Price index: Consumer prices (Mar 2021)