Change to Ukrainian interface versionChange to English interface versionChange to Russian interface versionHome pageClear last query resultsHelp page
Search for specific termsBrowse by subject categoryBrowse alphabetical list of titlesBrowse by organizationBrowse special topic issues

close this bookAlcohol-related Problems as an Obstacle to the Development of Human Capital (WB)
View the documentAcknowledgements
View the documentForeword
View the documentAbstract
View the documentIntroduction
close this folderThe nature of alcohol-related problems
View the documentHow do we measure alcohol consumption?
View the documentAlcohol-related problems
Open this folder and view contentsTrends in production and consumption
Open this folder and view contentsLevels and trends in alcohol-related mortality and morbidity
Open this folder and view contentsHow much do alcohol-related problems cost?
Open this folder and view contentsRole of government and policy options
View the documentConclusion
View the documentBibliography
View the documentAppendix tables
View the documentDistributors of world bank publications

How do we measure alcohol consumption?

Nobody knows exactly how much alcohol people drink. Although official statistics capture a portion of the consumption, in less developed countries a big share of intake is illicit home-brew which defies precise statistics. Estimates of black market production, importation, and sales for Latin America indicate that these consumption figures may underestimate true consumption by as much as one third'. Before entering into a discussion on the problems related to excessive alcohol consumption, it is useful to discuss how the consumption of alcohol is generally measured.

Ideally, consumption per capita would be measured through survey-based estimates, controlling for various socioeconomic characteristics which have been shown to influence alcohol consumption. Unfortunately, due to the dearth of information on actual consumption, these types of estimates are not generally available. In absence of true consumption data, the most commonly used proxy is an arithmetic mean of per capita consumption. Traditionally, finding the arithmetic mean would entail summing-up the alcohol consumption of a group of individuals and then dividing by the number of individuals in the group. Estimates of consumption in most countries, however, are generally based on information on the production, import and export, and sale of alcoholic beverages rather than alcohol demand studies. Thus, the figures used to describe consumption and over consumption really measure the availability of beverages legally on the market, rather than the absolute level of consumption. Despite these methodological problems, per capita consumption is widely accepted as a measure of alcohol consumption.

Once an average measure of consumption is decided upon, the next step is to gather evidence of heavy drinking. Although precise figures are not readily available, data on alcohol use/abuse are available from various small sample surveys. The distribution of alcohol consumption for sample populations in the United States, Costa Rica, and Mexico, age 18 to 65 years, is displayed in Table 1. Interestingly, the distribution of consumption in the three countries is similar. In the United States 33 percent of the population are classified as "abstainers" and 9 percent are "heavy drinkers, " while in Costa Rica 34 percent abstain and 10 percent are "heavy drinkers"; in Mexico 46 percent are "abstainers" and only 7 percent are classified as "heavy drinkers." In analyzing the distribution of alcohol consumption it is particularly important to examine the gender differences. In Mexico, for example, although only 7 percent of the population is classified as heavy drinkers, this is an average which reflects a 14 percent rate for the male population and only a 0.6 percent rate for the female population.

TABLE 1: Distribution of Alcohol Consumption

Amount Alcohol in ounces

Drinking Category

United States(%)

Costa Rica(%)







0.01-0 21

Light drinkers"-up todrinks weekly




0 22 0.99

"Moderate drinkers"- up to 2 drinks daily





"Heavy drinkers"- from 2-10 drinks daily





Source: US-Johnson (1977); Costa Rica -Miquéz (19831; Mexico Encuesta Nacional de Adicciones (1990).

In absence of survey data which reveal distributional drinking patterns and the incidence of cirrhosis of the liver, another commonly used proxy for excessive alcohol consumption is the mortality rate due to cirrhosis of the liver (Hymen, 1981). Although there is a lag between consumption and the emergence of the disease, cirrhosis mortality is a good indicator because the protracted consumption of alcohol in substantial quantities is toxic to the liver If consumption is sustained at a toxic rate, over time the cells of the liver degenerate and this protracted degeneration produces fibrosis in the liver. If the degeneration continues, death will likely result. Of course, not all cirrhosis mortality is the result of over-consumption of alcohol. While estimates of nonalcoholic cirrhosis fatalities range from 4 to 50 percent, the literature indicates that 50 percent of all cases of cirrhosis are related to alcohol (Hymen, MM, 1981). In addition',it is important to note that only 10 to 25 percent of clinical alcoholics have well-stablished cirrhosis of the liver, apparently due to the long lag between the onset of heavy drinking and the manifestation of cirrhosis in the liver.

Another measure of excessive consumption, which is generally available in national mortality statistics, is the number of deaths attributable to alcohol dependence syndrome. Alcohol dependence syndrome is a clinical classification of alcoholism adopted by the International Classification of Diseases (ICD-10). Given that nearly 100 percent of the deaths under this classification can be directly attributable to a history of excessive alcohol consumption, this may, in fact, be a better proxy for excessive consumption than any of the other proposed indices By analyzing mortality rates for either cirrhosis of the liver or alcohol dependence syndrome concurrently with per capita levels of absolute alcohol consumption, a general picture of the range and severity of alcohol-related problems in a country can be ascertained.

The influence of alcohol substance abuse on morbidity and the subsequent link to declines in productivity are other important aspects in analyzing alcohol-related problems. Special studies designed to measure the percentage of hospitalizations or emergency room cases which are alcohol related are generally the most effective measure; though detailed studies of the discharge diagnoses of patients admired for alcoholrelated problems provide a much more precise figure of the prevalence of alcohol abuse in the population. Evidence from several studies based on hospital roster-admissions of morbidity in developing countries will be presented later in this section.

to previous section to next section

[Ukrainian]  [English]  [Russian]