Monthly Survey of Manufacturing Industry (EMIM) 2018

The Monthly Survey of the Manufacturing Industry (EMIM) emerged in 1964, which makes it one of the longest-running surveys of economic units in our country today.

In order to reflect the economic reality of the national manufacturing industry, during its 60 years of experience, the EMIM has guaranteed the representativeness of the statistics produced. In this context, said survey has been updated periodically as described below.

• In its beginnings, information on 29 classes of economic activity were collected, processed and disseminated, taking a total of 604 establishments in the manufacturing sector as a sample. The Mexican Catalog of Economic Activities (CMAE) was used as a classifier.
• In 1976, the survey increased its measurement to 57 classes of economic activity of the CMAE and 1,338 establishments under study.
• In 1987, new classes of activity were incorporated, reaching 129, which implied considering 3,218 establishments in the sample. The CMAE is still used as a classifier.
• In 1994, 205 classes of economic activity and 6,867 establishments were reached. The Mexican Classification of Activities and Products (CMAP) version 1994 is used.
• It was in 2005, when in addition to increasing the number of classes of economic activity under study to 230 and forming a sample with 7,238 establishments, the North American Industry Classification System (NAICS) version 2002 was implemented for the first time.
• In 2007, the number of classes of economic activity under study increased to 240 and the number of establishments increased to 11,406, applying the NAICS version 2007.
• In 2013, the EAIM started a statistical series with 10,447 establishments and the number of classes of economic activity to 239, also applying the North American Industry Classification System (NAICS) 2013.

In 2018 the current EAIM series begins. In this new series, 206 types of activity are measured according to NAICS 2018 and there is a sample made up of 8,819 economic units, which makes it possible for us to reflect the activity of the Mexican manufacturing sector in accordance with the prevailing economic reality in the country.

An outstanding event in the long history of the EMIM was determined on December 14th, 2012, when the Governing Board of the INEGI determined this survey as information of National Interest, by fully complying with the four criteria established in article 78 of the Law of the National System of Statistical and Geographical Information.

Target population

These are the establishments classified within the 206 classes of economic activity of interest belonging to sector 31-33 Manufacturing Industries, according to the North American Industry Classification System (NAICS) 2018.

Thematic coverage

1. Days worked.
2. Personnel dependent on the company name (number of people).
3. Personnel not dependent on the company name (number of people).
4. Hours worked by the personnel dependent on the company name (thousands of hours).
5. Hours worked by the personnel not dependent on the company name (thousands of hours).
6. Remunerations paid to personnel dependent on the company name.
7. Percentage of plant capacity used.
8. Expenses for consumption of goods and services (national and foreign market).
9. Income from the supply of goods and services (national and foreign market).
10. Production value of the manufactured products by product.
11. Sales value of the manufactured products by product.
12. Export sales of manufactured products by product.

Classifier used 

North American Industry Classification System (NAICS),  version 2018.

https://www.inegi.org.mx/app/biblioteca/ficha.html?upc=702825099695

International recommendations
International Recommendations for Industrial Statistics 2008 of the United Nations.


National recommendations

INEGI standards.

• Standard for Quality Assurance of Statistical and Geographical Information.
• Technical Standard for the Production Process of Statistical and Geographical Information for the National Institute of Statistics and Geography.
• Guidelines for updating the economic information generated by the General Directorate of Economic Statistics. 

SNIEG Standards.

• Agreement for the use of the North American Industry Classification System (NAICS) in the collection, analysis and presentation of economic statistics.
• Principles and Good Practices for Statistical and Geographical Activities.
• Technical Standard for the Generation of Basic Statistics.

Sample size

8,819 establishments (in design year 2018)

Observation unit

Manufacturing establishment

Sampling frame

The statistical framework of reference is the Industrial Census that is carried out in Mexico regularly every five years, therefore the 2019 census directory is the population framework from which the main classes of activity and the establishments in the sample were selected. The selection criteria applied are defined by the value of the income reported by the establishments in said census.

Sampling scheme

Of the total of 206 domains, a non-probabilistic design is defined for 181 and a probabilistic design for the remaining 25 domains.

The conduction period takes place from April to July of the year following the study period, so on average it lasts four months.

The EMIM collection instrument and the information is obtained under the following collection modalities.

• Printed questionnaire (on paper).
• Electronic questionnaire for collection on mobile computing devices.
• Electronic questionnaire on the INEGI website on the Internet (direct collection within the application).
• Electronic questionnaire on the INEGI Internet website (Excel format for downloading, independent filling, and subsequent uploading to the site).

When using microdata from surveys with probabilistic designs for purposes of analysis and research of economic phenomena, it is pertinent to bear in mind the considerations mentioned below: 

1. A probabilistic sample is selected according to the specific sampling scheme used in the design of the particular survey to be carried out and depends on the universe to be investigated, the variables that are intended to be collected, the information that needs to be estimated and the accuracy desired in the estimates to be made. 
   

    

2. The information estimated by a probabilistic survey results from a process of applying the corresponding expansion factors to the microdata obtained. The expansion factor is a number calculated based on the characteristics of the sampling scheme used and that allows the values obtained from the sample to resemble the values of the population under study. Technically the expansion factor is the inverse of the probability of selection of each economic unit considered in the sample conducted. 
   

    

3. To make good use of microdata, it is recommended to comply with the specifications set out in the sampling design. Therefore, it is recommended to consult the methodological document where it is explained in detail how the sample was designed and what the study domains are. 
   

    

4. Depending on each probabilistic survey, the design of the sample could cover one or several study domains separately, or a combination of them at the same time, this according to the purposes and scope of the statistical research to be carried out. Some examples of these may be the following: 
   

    

   • National
   • By state
   • City
   • By sector, subsector, branch or class of activity
   • Urban or rural
   • By size 

    

    
   The information estimated with levels of precision and confidence in accordance with the characteristics embodied in the sample design, can only be done respecting the study domains of the survey.
   

    

5. It is recommended not to estimate information for the variables, spatial segmentations or subgroups of classification of economic activity for which the statistical design of the survey was not performed. For these, the information should be handled only at the sample level (without expanding) and used to identify trends, structures or behaviors of the phenomenon to be analyzed.

In the case of surveys with non-probabilistic design, the following should be considered: 

1. A non-probabilistic sample (deterministic) consists of selecting the units under study according to their percentage contribution in the design variable that can be: total income, production value or personnel employed until reaching the coverage established for each study domain.
   

    

2. Considering the specific characteristics of each study domain, the selection criteria are defined according to the type of coverage that is intended to be achieved in the selection or design variable, for example: 

High: It corresponds to the study domains whose coverage is greater than or equal to 80% of the design variable.
Medium: It corresponds to the domains whose coverage is greater than or equal to 60% but less than 80% of the design variable. 
Low: It corresponds to the domains whose coverage is less than 60% of the selection variable. 

• In a survey with non-probabilistic design (deterministic) , the data to be analyzed are the samples, whose results denote the behavior or evolution of the study domains. Therefore, expansion factors are not calculated or applied. 

Uses of information from the Monthly Survey of the Manufacturing Industry (EMIM)

1. If the analysis to be carried out with the microdata comes from the study domains of the EMIM (class of economic activity) where a probabilistic design is applied, this analysis can be carried out at the national level, sector, subsector, branch and class of activity, where it is possible to use the estimated information obtained by applying the expansion factor to the sampling value of the microdata. In this analytical context, the work of statistical, mathematical or econometric modeling carried out through the microdata laboratory or through remote information processing is framed. 
   

    

2. When the data user with probabilistic design wishes to carry out an analysis or research on a specific set of microdata that is not framed within the study domains of the survey, or due to a special consideration or restriction of the desired analysis, the microdata at a level more disaggregated than that defined in the study domain, then only the sampling data should be used. That is, within the Microdata Laboratory or through remote information processing, only the sampling data that is not expanded will be used. If the researcher decides to apply the expansion factor to this specific subset of data,the INEGI will not be held responsible for the estimates obtained, because they do not correspond to the study domains for which the sample was designed. 
   

    

3. If the analysis to be carried out with the microdata comes from domains with non-probabilistic design (deterministic) of the EMIM, this can be at sector, subsector, branch and class of activity at the national level; any different geographical disaggregation does not represent the survey and the INEGI will not be responsible for these calculations.
   

    

4. In the case of the use of the EMIM data from domains with different design that implies a combined analysis; that is, both the probabilistic and non-probabilistic design (deterministic), it is suggested that the analysis of the microdata will be at the national level, sector, subsector, branch and class of activity, and should be subject to the characteristics already specified in each case.