Instituto Brasileiro de Geografia e Estatística

A- A+

System of Consumer Price Indices

Technical Note - 01/2001

Alteration in the Collection Methodology of Domestic Personnel

  • From this year on, in order to estimate the indices produced through the the National System of Consumer Price Indices (SNIPC), “domestic personnel” will be estimated based on the incomes collected by IBGE’S Monthly Employment Survey – PME. PME is one of the main sources of labor statistics. It is a monthly household survey that produces several statistics about the structure/distribution of the economically active population, employment/unemployment levels, average incomes of the employed population, etc.
  • Up to 2000 the national minimum wage was used in the estimate. However, some studies showed that in the last years the average of incomes of the domestic personnel did not accompany the change of the minimum wage. They also revealed that PME is an appropriate source, since it investigates the incomes effectively earned by the domestic personnel.

In this note IBGE explains the new methodology.

Definition of Domestic Personnel

In order to identify a homogeneous group of professionals, only the monthly income of the domestic workers with the following characteristics were considered:

  • Non-specialized, who performs any task in the housing unit. Not applied to those who perform specific functions such as babysitters, gardeners, etc.
  • Genuine employees, with a formal contract or not. "Own-account"workers are excluded.
  • Those receiving wages monthly, every 15 days or weekly.

Method of calculus

  • The change resulting from the comparison of the estimate of the the average income series trend in the T month with the average income series trend trend in the T-1 month is considered every month. The trend of a variable is the statistics derived from its original series relative to a long period of time. The general direction of this series is indicated.

  • The option for the trend is determined by the presence of specific fluctuations in the original series, of which causes are not predicted by the theory of consumer price indices. Extraordinary earnings such as the thirteenth salary (a bonus equivalent to one month’s salary) or the vacation benefit (a bonus of one third of one month’s pay) are examples of fluctuations in the series that are not predicted by the theory.

  • The trend series will be obtained by the application of the X-12 ARIMA system to the original series of average incomes, considering January of 1995 as the starting point and the last month informed as the final one. The trend observed in the last month will be compared with the trend of the previous month, similarly to the method of calculus of monthly indices.

Geographical coverage

  • The result of the trend of each one of the six areas surveyed by PME will be considered in the corresponding area of SNIPC. Thus, Rio de Janeiro, Porto Alegre, Belo Horizonte, Recife, São Paulo and Salvador will express the characteristics of their local labor markets. The average income series trend obtained from all the information of the six areas surveyed by PME will be attributed to the other five areas of SNIPC not covered by the survey (Brasília, Belém, Fortaleza, Curitiba and Goiânia).

Reference month

  • The reference for the income change estimates of domestic personnel for a T month is all the information obtained by PME in a T-2 month, of which the surveyed incomes refer to the T-3 month. In other words, given a price index of January, the result of domestic personnel will be obtained from the PME of November of the previous year, when the incomes relative to October of the previous year were surveyed. By the end of every year the price indices will have incorporated the accumulated change of 12 months of PME, from November of the previous year to October of the present year.
  • Specifically in this year, the accumulated index obtained from the PME of four months (November of 2000 to February of 2001) will be incorporated in the indices of April. From May on, the indices will reflect results relative to one month, starting with March of 2001.