Defining Socio-economic Levels in Latin America

In the traditional analysis of the media audiences in Latin America, the three most important variables are age, sex and socio-economic level. Many media are targeted towards specific age/sex groups (e.g. cartoons for children, sport programs for men, telenovelas for women, etc.), so that the importance of identifying target groups in terms of age/sex is self-evident. Socio-economic level is also important, because of the significant differences in spending power among different socio-economic levels. For many luxury consumer goods, the target audience is likely to be only from the upper socio-economic strata.

Whereas the definitions of age and sex are unambiguous, socio-economic level is much more difficult to articulate. In the USA, it is customary to use annual household income (before taxes and including wages, tips, alimony, interests and other sources of income) as the single variable that would best encapsulate spending power. This is possible through a unique set of circumstances that do not necessarily exist elsewhere (for example, income taxes are filed each year so that most people in the USA have at least an approximate idea of their 'reported' household incomes). In Latin America, household income is problematic due to a variety of reasons:- hyper-inflation, non-cash and barter activities, unsteady cash flows, absence of household financial bookkeeping, differences in costs of goods and services, as well as a general reluctance to reveal household wealth.

In a number of Latin American countries, the definition of socio-economic level was formulated under the aegis of the national marketing associations. Typically, the definition is given in terms of a point-scoring algorithm. Here are some examples:

Argentina: (La Asociación Argentina de Marketing) This is a point-scoring scheme based upon the educational achievement and occupation of the main income earner, the possession of consumer goods (color television with remote control, VCR, refrigerator with freezer, freestanding freezer, personal computer, telephone, automatic washing machine, clothes dryer, credit card, air conditioner), the possession of automobile (number of cars, year, make, model) and housing characteristics (architectural style, size, materials and appearance).

Brazil: (ABA/ANEP) This is a point-scoring scheme based upon the number of television sets, radio sets, automobiles, domestic helpers, vacuum cleaners, washing machines and bathrooms, and the educational achievement of the head of household.

Mexico: (La Asociación Mexicana de Agencias de Investigación de Mercado y Opinión Pública) This is a point-scoring scheme based upon the educational achievement and occupation of the head of household, the ownership of goods (water heater/boiler, toaster, vacuum cleaner), the number of light bulbs in the home, the number of rooms (excluding bathrooms) and the number of servants.

Typically, these definitions were obtained through a careful examination of a large number of variables and selecting just those essential, non-redundant variables that convey the most information about socio-economic well-being. The approach is necessarily reductionistic, since the objective is to ask as few questions as necessary to determine the socio-economic level. In each country, the selected variables are those that best reflect the social and economic realities there. Consequently, it is expected that the definitions would be different across countries.

In the Los Medios y Mercados de Latinoamérica study, we needed to come up with a pan-regional definition of socio-economic level. We could not use the national definition in each country, since the results would not be comparable across countries. We could not use the definition of one country (say, Argentina) everywhere, because there might be unique circumstances. For example, air conditioners are not needed in high-altitude places such as La Paz, Bolivia.

For the 1995 edition of Los Medios y Mercados de Latinoamerica study, we used the following algorithm:

With these scores, we classify the households into socio-economic groups. The top 10% is called the level "A", the next 20% is called level "B", the next 30% is called level "C" and the bottom 40% is called level "D".

Our approach is diametrically opposite to those used for the different national definitions because we operate under different constraints. For the national marketers, they needed to determine socio-economic level accurately with as few questions as possible. Therefore, their approaches may be characterized as being reductionistic.

For us, the opposite is true. In our study, we had already collected all sorts of demographic and product consumption and ownership data. We can afford to use as many of the collected variables as we feel like. This allows us to have a robust definition that would minimize the impact of the breakdown of a single variable in some countries.

In the next table, we characterize the social and economic characteristics of our four levels.

  Level A Level B Level C Level D
HH income US$24,000+ 34% 9% 3% 1%
HH income US$36,000+ 20% 2% 1% 1%
College Graduate 17% 7% 5% 1%
Read written English 16% 4% 1% 0%
Understand spoken English 15% 3% 1% 0%

(Source: Los Medios y Mercados de Latinoamérica 1995)
(Explanation: 16% of the Level A people can read English, 4% of the Level B people do so, and so on).

The data in the above table match our expectations about the relationship between our composite socio-economic variable and some traditional direct measures of economic well-being and educational achievement.

Being media researchers, we are interested in the relationship between media behavior and socio-economic level. This is summarized in the following table:

  Level A Level B Level C Level D
Multichannel Television* 45% 30% 20% 9%
Read weekday newspaper 49% 37% 25% 15%
Read regional magazine** 56% 40% 34% 17%

(Source: Los Medios y Mercados de Latinoamérica 1995)
(* Multichannel television = Direct-To-Household satellite TV, Satellite Master Antenna Television (SMATV), wired cable television, wireless cable television (aka MMDS))
(** A regional magazine is a magazine that is distributed in several countries. Examples are Reader's Digest, Time, Newsweek, National Geographic, América Economia, TV y Novelas, Vanidades, Buenhogar, etc.)
(Explanation: 49% of Level A persons read newspapers on weekdays, 37% of Level B persons do so, etc.)

In the two tables above, we have established two simple facts:

These two facts would suggest the multichannel television and print media are efficient media to reach the upper socio-economic groups.


(posted by Roland Soong on 1/11/97)

(Return to Zona Latina's Home Page)