Alcohol consumption and poverty in Sri Lanka

Scientific Study 2017 63 Pages

Sociology - Miscellaneous



1 Introduction

2 Alcohol Production and Consumption
Types of alcoholic beverages
Global alcohol production and consumption

3 Alcohol Production and Consumption in Sri Lanka
Concepts of poverty and its definitions
Measurements of poverty

4 Global Poverty

5 Poverty in Sri Lanka

6 Relationship between Alcohol Consumption and Poverty

7 Concluding Remarks


Alcohol consumption and poverty in Sri Lanka

Discussion paper

Dominic Williams[1]

1 Introduction

This Paper discusses a detailed description of the common types of alcohol and a global overview of alcohol consumption and poverty. It will also present a detailed discussion about alcohol production and consumption patterns as well as poverty in the country of interest of this thesis, Sri Lanka.

The structure of this chapter is as follows. Section 2 provides a global overview of alcohol production and consumption. The next section presents details about the Sri Lankan alcohol industry and alcohol consumption patterns. Section 4 presents a global overview of poverty in various countries around the world and the next section describes poverty in Sri Lanka. Section 6 briefly provides an overview of the relationship between alcohol consumption and poverty. The final section, of this chapter, presents the concluding comments.

2 Alcohol Production and Consumption

Types of alcoholic beverages

Alcoholic beverages come in different varieties with various strengths of alcohol content and taste. The most commonly consumed alcoholic beverages globally are beer, wine and (distilled) spirits. Distilled spirits include beverages such as brandy, whisky and rum, and are generally higher in pure alcohol content than that of beer or wine. There are also a number of other country-specific alcoholic beverages, which include, for example, sake originating from Japan, vodka from Russia, and arrack and toddy produced mainly in South Asian and South-East Asian countries. Kasippu is another type of alcohol, high in pure alcohol content, and is produced illegally in some of the South-East Asian countries, especially in Sri Lanka. Table 1 presents information on the various commonly consumed alcoholic beverages, raw materials used to produce them and the percentage of their pure alcohol content.

Table 1: Types of alcoholic beverages, their source and strength

illustration not visible in this excerpt

Source: Based on WHO (2004).

As shown in the table 1, the different types of alcoholic beverages are derived from different fermenting ingredients. As can also be seen, beer, toddy and wine contain the least pure alcohol content while the others contain much higher alcohol content. It should be noted that the consumption of alcohol is legal in most countries, while consumption of non-licensed beverages such as kasippu is illegal.

Global alcohol production and consumption

Globally, over the past three decades, alcohol has become much more readily available. Although some cultures may have placed formal bans on alcohol consumption, consumption of alcohol is very much part and parcel of human lives in many countries. According to a World Health Organisation report WHO (2011), the world’s 2005 average consumption of total alcohol was 6.13 litres per capita.

The following two tables, Tables 2 and 3, are drawn from the Food and Agricultural Organisation (FAO) and the United Nations Statistical Office statistical databases for the production of beer, wine, spirits and fragmented beverages by country. The FAO classification for ‘beer’ includes malt beers; ‘wine’ includes wine made from grapes; ‘spirits’ include all distilled beverages; and ‘other’ includes other fermented beverages

Table 2: Global production of beer, wine, spirits, other and total alcohol, 1961-2009 (millions of tonnes)

illustration not visible in this excerpt

Source: Based on FAO (2013) data base, August 2013.

Table 3: Growth in global production, beer, wine, spirits, other and total alcohol, 1961-2009 (in percentages)

illustration not visible in this excerpt

Source: Based on FAO (2013) data base, August 2013.

made from sorghum, maize, millet, rice, cider, fruit wine, fortified wine, and the like. Total alcohol covers beer, wine, spirits and other alcohol beverages. Tables 2 and 3 present the global production[2] and growth (in percentages), respectively, of beer, wine, spirits, other beverages, and total pure alcohol from 1961 to 2009. As can be seen, world production of total pure alcohol grew from about 80 million tonnes, in 1961, to 237 million tonnes, in 2009, equating to a total increase of 195 percent for this period. This represents an annual average growth rate of 3 percent over the sample period. For this same period, global population has only increased from 20 billion to 46 billion, an annual average growth rate of 1.6 percent. Naturally, demand for alcoholic beverages and, hence, the production will grow as population grows; but, alcohol production has grown at a much faster rate than the population growth for the same period.

Considering the production and growth of individual beverages in Tables 2 and 3, global production of beer, between 1961 and 2009, has expanded from about 43 million tonnes to 167 million tonnes, which represents an average growth rate of 2.87 percent annually. In contrast, production of wine rose from about 21 million tonnes in 1961 to 23 million tonnes in 2009, giving an overall modest 0.24 percent average annual growth rate. It can also be seen that global spirits production has increased at an average annual growth rate of 2.4 percent; growing from about 7 million tonnes in 1961 to 20 million tonnes in 2009. Production of other alcoholic beverages has also grown at an annual average rate of 2.33 percent. Figure 1 presents the global production of beer, wine, spirits, other and total alcohol for the period 1961 to 2009. It can be seen that, while production of beer has risen at a very fast rate than all other beverages. The production of wine has increased at a very low rate until 1984 and has then declined slightly or stayed the same. It is worth noting that wine production depends on the availability of land for cultivation of grapes and the weather condition. The production of spirits and other alcoholic beverages have also remained stable over the 49-year period. Looking at the total alcohol production globally, it can be seen that this has increased at the same rate as beer. That is, the growth in alcohol production is solely explained by the growth in beer production.

Figure 1: Global production of alcohol, 1961-2009

illustration not visible in this excerpt

Source: Based on FAO (2013) data base, August 2013.

Figure 2 displays the shares of beer, wine, spirits and other alcoholic beverages production for 1961 to 2009, obtained from Table 2. As can be seen, in 1961, beer held 53.9 percent of total alcohol market share; wine’s share was half that of beer at 26 percent. Other alcoholic beverages held 11.5 percent while spirits’ share was 8.6 percent. However, by 2009, the share of beer production has risen by almost 16.4 percent, (from 53.9 percent to 70.3 percent) but the share of wine has declined by about the same percentage with a fall of 16.7 percent (from 26.0 percent to 9.3 percent). The share of spirits and other alcoholic beverages have remained the same, around 8.6 percent and 11.5 percent, respectively. In other words, beer production has increased at the expense of wine, implying that global consumers are moving away from relatively higher alcohol content beverages (wine) to lower alcohol content beverages (beer). This could be due to a number of governments’ sponsored anti-alcohol drink-driving and other campaigns[3]. and per adult beer, wine, spirits and total alcohol consumption in litres of pure alcohol in 2005 for 188 countries published by WHO (2011). Countries are listed based on descending order of GDP per capita. As can be seen from the GDP column (4) and the total alcohol consumption column (8), in general, countries with higher per capita income has higher level of alcohol consumption, except for Arabic countries, where official ban, for religious reasons, on alcohol consumption is enforced. Most African and Asian countries have lower per capita income and also consume relatively low volume of pure alcohol. However, it is worth noting that poor countries tend to have more illegal alcohol producers and these tend to cost less and are very attractive to poor consumers .

Figure 2 : Global shares of alcohol production, 1961-2009

illustration not visible in this excerpt

Source: Based on data base, August 2013.

Table 4 presents the Purchasing Power Parity (PPP) adjusted Gross Domestic Product (GDP) per capita

Table 5 presents the top 25 countries in terms of per adult beer, wine, spirits and total alcohol consumption (in litres of pure alcohol) and their ranks in 2005. (The corresponding full table for all 188 countries is given in Table A1 of Appendix A2.) For example, Australia is ranked 21st in terms of beer and 25th in terms of wine consumption. Looking at the total alcohol consumption, Moldova (18.22 litres) is ranked first, followed by the Czech Republic (16.45 litres) and then Hungary (16.27 litres).

In terms of per adult alcohol consumption of beer, Palau (8.68 litres) is ranked as number one followed by the Czech Republic (8.51 litres) and Seychelles (7.15 litres). With respect to wine, Luxembourg (8.16 litres) is ranked as number one followed by France (8.14 litres) and Portugal (6.65 litres). For spirits, Korea (9.57) is ranked as number one followed by Estonia (9.19 litres) and Saint Lucia (8.21 litres). In 2005, in pure alcohol, average per adult consumption of beer was 1.96 litres, wine was 1.03 litres, spirits was 1.64 litres, and total alcohol was 6.52 litres. A closer look at the top 25 ranked countries for beer and wine shows that most of them are European countries. Top 25 spirits consuming countries is mixed among the various continents. In terms of total alcohol consumption, most of the top 25 countries are from Europe.

Table 4: GDP per capita and per adult beer, wine, spirits and total alcohol consumption (in litres of pure alcohol), 188 countries, 2005

illustration not visible in this excerpt

(Table continues on the next page.)

Table 4: GDP per capita and per adult beer, wine, spirits and total alcohol consumption (in litres of pure alcohol), 188 countries, 2005 (ranked in order of descending GDP) (continued)

illustration not visible in this excerpt

(Table continues on the next page.)

Table 4: GDP per capita and per adult beer, wine, spirits and total alcohol consumption (in litres of pure alcohol), 188 countries, 2005 (ranked in order of descending GDP) (continued)

illustration not visible in this excerpt

Note : n.a denotes data that are not available.

Source: and GDP figures extracted form (http://data.worldbank.org/topic), and GDP for * noted countries extracted from http://www.indexmundi.com, accessed on August 26, 2013.

Table 5: Top 25 countries with the highest per adult consumption of beer, wine and spirits and total alcohol (in litres of pure alcohol), 2005

illustration not visible in this excerpt

Source: Based on WHO

Table 6[4] presents the average per adult consumption of beer, wine, spirits, other alcoholic beverages and total alcohol (in litres of pure alcohol) for each continent. The continents are ranked based on their average per capita GDP (in $US, PPP adjusted). The last row presents the average world consumption. In terms of pure alcohol, world alcohol consumers consume more beer than any other alcoholic beverage, followed by spirits, wine and then other country-specific beverages. On average, per capita consumption of total alcohol was 7.23 litres of pure alcohol in 2005. Figure 3 presents the corresponding bar chart for each continent. Overall, it can be seen from Table 6 and Figure 3 that, continents with higher income per capita tend to consume more alcoholic beverages than those with lower income. African countries recorded the lowest per capita total alcohol consumption followed closely by the Asian and Oceania countries. European countries have the highest average GDP and recorded the highest average consumption of alcohol which is almost twice as that of most other continents. Looking at the different continents, Africans consume more of beer and other beverages compared to wine and spirits, while Asians consume more of spirits than the other three beverages. Oceania consumers consume more beer than the other alcoholic beverages. North Americans consume mostly spirits followed by beer. Europeans consume mostly beer, wine and spirits.

Table 6: Average GDP and per adult alcohol consumption, 6 continents, 2005

illustration not visible in this excerpt

Source: Based on database, August 2013.

Figure 3: Average per adult consumption of beer, wine, spirits and other alcoholic beverages (in litres of pure alcohol), 6 continents, 2005

Abbildung in dieser Leseprobe nicht enthalten

Source: Based on data base, August 2013.

Figure 4 presents the per adult consumption for the continents by alcoholic beverages. On average, the Europeans are the highest beer and wine consumers while Asians consume the least of both beverages. North Americans consume the most spirits followed by the Europeans and then the Asians, while Africans consume the least. Africans consume mainly beverages other than beer, wine and spirits.

Figure 4 : Average per adult consumption in the six continents by alcoholic beverages (in litres of pure alcohol), 2005

illustration not visible in this excerpt

Source: Based on FAO (2013) data base, August 2013.

Next we group the countries into less developed countries (LDC) and developed countries (DC). Figure 5 presents the scatter plots of alcohol consumption against GDP per capita for developed countries (DC) and less-developed countries (LDC)[5], using data from columns (8) and (4), Table 4. As can be seen, there is a positive relationship between GDP per capita and alcohol consumption for both sets of countries. It can be interpreted that, for every $1000 increase in GDP per capita, per capita alcohol consumption would increase by 0.018 litres of pure alcohol in the DCs, whereas it would increase by 0.451 litres of pure alcohol in the LDCs. It should also be noted that the slope is significant for the LDCs (p -value = 0.00) and insignificant for the DCs (p -value = 0.78). This shows that alcohol consumption is much more sensitive to income changes in poor countries than in the rich countries.

Figure 6 illustrates the average alcohol consumption of consumers in the LDC and DC. In general, consumption of alcohol is much higher in the developed countries than in the less developed countries. As can be seen, consumption of beer and wine are highly dominated by consumers from developed countries while spirits consumption is, comparatively, similar in all countries. Other country-based alcoholic beverages are consumed more in the less developed countries than the developed countries.

Figure 5 : Global Shares of Alcohol Production, 1961-2009

illustration not visible in this excerpt

Source: Based on FAO (2013) and World Bank (2013a)

Figure 2.6: Average per capita beer, wine, spirits, other and total alcohol consumption (in litres of pure alcohol), LDC and DC, 2005

illustration not visible in this excerpt

Source: Based on FAO (2013) date base, August 2013

Figure 2.7: Relationship between per capita alcohol consumption and per capita GDP, LDC and DC, 2005

illustration not visible in this excerpt

Source: Based on FAO (2013) and World Bank (2013a)

Figure 7 explores further the relationship between alcohol consumption and GDP per capita for less developed and developed countries. It is evident that in the LDCs, as GDP increases, consumption of beer, wine and spirits also increases but the consumption of other beverages decrease. That is, the poorer LDCs tend to consume more of other beverages with high alcohol content. Within the developed countries, GDP doesn't seem to have much effect on the consumption of beer, wine, spirits or total alcohol. In conclusion, Figure 7 highlights the fact that the level of income has a significant effect on the consumption of alcoholic beverages in the LDCs, but it is not the case with the DCs. That is, consumption patterns of the different alcoholic beverages, in relation to the level of income, tends to be dissimilar between the less developed and developed counties.

This section provided a global picture on the consumption of beer, wine, spirits and other alcoholic beverages. In the next section, this study will focus on the Sri Lankan case, analysing the consumption patterns of the various legal alcoholic beverages available in that country.

3 Alcohol Production and Consumption in Sri Lanka

The Sri Lankan Alcohol Industry

In modern Sri Lanka, the manufacture of legal alcohol dates back to 1913. The Excise Department was constituted to manufacture a particular liquor, arrack, for the local market. In 1974, the State Distilleries Corporation was formed to take over the distilling and bottling activities of the said department; the Excise Department continued to function as the monitoring and controlling authority of the alcohol industry. From 1913, until the Distilleries Corporation of Sri Lanka was privatized in 1992, the government of Sri Lanka had virtual monopoly on the production and distribution of the alcoholic beverages.

As highlighted in Section 2, in addition to the mostly imported variety of beer, wine and spirits, consumers in different countries also consume various locally produced alcoholic beverages. In Sri Lanka, alcoholic beverages are widely and freely available. These beverages consist, mainly, of locally manufactured legal alcoholic beverages, for example, arrack and toddy; imported foreign alcoholic beverages, such as whisky, brandy, gin, vodka and rum; and locally produced and imported beer and wine. The 2012 administration report of the Excise General Commissioner listed 72 manufacturing institutions and distillery companies as being legally involved in the production of liquor. Amongst these, 20 companies produce arrack and 26 produce spirits. Moreover, 8 companies produced country made foreign liquor. In addition, 32 toddy manufacturing companies were operating in 2012. Only two local companies produce only beer and wine. (See Table A2, Appendix A2 for more details).

Since the privatization of Sri Lanka’s alcohol industry in March 1992, the volume of liquor produced in the country has increased rapidly. Liquor production is usually measured in ‘proof’ litres. Table 7 and Figure 8 presents the Sri Lankan alcohol production of low and high alcohol content beverages and the corresponding government revenue[6] for the period 2001 to 2012. There has been a sudden increase in the production of all alcoholic beverages, except spirits, in 2009 and continued until 2012. This could possibly be due to the removal of obstacles for liquor sales in the Northern and Eastern areas of the country and the significant increase in foreign tourist arrivals with the end of the war in 2009. In 2001, total alcohol production was 70 million proof litres; this volume increased to 147.1 million proof litres in 2011. This represents an annual average rate of increase of 8 percent. Taking into consideration the low alcohol content beverages, in 2001, the production of beer and toddy was recorded as 40 million proof litres and 2.6 million proof litres, respectively, with beer more than doubling to 99.3 million proof litres and toddy doubling to 5.8 million proof litres by 2012. In 2001, local wine production was 0.022 million proof litres but drastically decreased to 0.007 million proof litres in 2008 and then increased to 0.036 thousand proof litres in 2012.

From Table 7, with respect to the high alcohol content beverages, arrack production in year 2001 was 25.7 million proof litres and nearly doubling to 45.2 million litres over the sample period. Liqueur production rose from 1.1 million to 4.2 million proof litres but spirits production fell from a tiny volume of 0.113 million proof litres to nil production.

From Figure 8, which presents the production of the alcoholic beverages graphically, it is evident that beer, aside from having the highest production, also has the highest growth rate in production. Arrack comes second in production volume as well as in production growth rate.

Table 7: Quantitya from manufactured alcoholic beverages, 2001-2011, Sri Lanka

illustration not visible in this excerpt

Note: a in millions of proof litres; b includes brandy, gin, rum, whisky.

Source: Administrative Report of the Commissioner General of Excise, various issues.

Figure 8: Production of alcoholic beverages, Sri Lanka, 2001-2011

illustration not visible in this excerpt

Source: Administrative Report of the Commissioner General of Excise, various issues.

Table 8 and Figure 9 present the government tax revenue[7] from alcoholic beverages. Overall, government tax revenue from alcoholic beverages has increased from around Rs. 6 billion in 2001 to 28 billion in 2009, an annual growth rate of 20 percent. The sudden increase in the production of all alcoholic beverages in 2009 has led to the Sri Lankan government receiving tax revenue at a higher rate. For example, from 2010 to 2011 total alcohol tax revenue from alcoholic beverages has increased to Rs. 56 billion from Rs. 38 billion; a windfall of over 40 percent annual average growth rate. This sudden increase in total alcohol revenue could be due to two possible reasons. Firstly, business turn-over tax was abolished; instead it was added to the excise tax, and the Sri Lankan government raised its excise tax rates twice in 2011 (S. Fernando, 2017).

As can be seen from Table 8 and Figure 9 the largest amount of revenue has been collected from arrack. This represents more than 75 percent of the total revenue for every year of the sample period. In 2001, arrack revenue was Rs. 5 billion and grew at an annual average rate of approximately 22 percent to Rs. 44 billion in 2012. The second largest amount of tax income has been earned from duty on beer; 785 million rupees was collected in 2001 and soared to Rs. 10.4 billion by the year 2012. The revenue from these two beverages increased by an annual average rate of around 25 percent during the period 2001 to 2012. However, after 2007, there was a considerable revenue increase from toddy. This was due to the new e xcise duty imposed on the bottled toddy manufacturing in 2007; at a rate of Rs 10 per litre on bottled toddy. Furthermore, spirits is the alcoholic beverage that the Sri Lankan government has imposed the least amount of excise. Thus, this category has provided the government with very little revenue; this revenue has decreased gradually over the years and stopped when production ceased after 2008.Table 2.8: Revenue from manufactured alcoholic beverages, 2001-2011, Sri Lanka

illustration not visible in this excerpt

Note: a in Rs. millions; b includes brandy, gin, rum, whisky.

Source: Administrative Report of the Commissioner General of Excise, various issues.

Figure 2.9: Government tax revenue from alcoholic beverages, Sri Lanka, 2001-2011

illustration not visible in this excerpt

Source: Administrative Report of the Commissioner General of Excise, various issues.

Legal and Illegal alcohol ic beverages

Tables 9 and 10 present the estimated total consumption and per capita consumption of legal alcoholic beverages, for the period 2000 to 2007. The data has been drawn from the Sri Lankan Excise Commissioner’s Annual Reports. Total consumption of arrack and beer grew from about 100 million litres, in 2000, to 120 million litres in 2007. Consumption of legally produced bottle toddy was 5.6 million in 2000 and almost doubled by 2007. As can be seen in Table 10, total per capita consumption of pure alcohol in Sri Lanka also grew rapidly over the same period. In 2000, total per capita pure alcohol consumption was 1.76 litres and increased to 21 litres in 2007. Furthermore,, there have been significant changes in the different types of per capita consumption beverages. Between 2000 and 2007, per capita consumption of total alcohol increased by 25 percent, with the per capita consumption of foreign liquor increasing by 100 per cent. Per capita consumption of arrack had increased by almost one-fifth while the per capita consumption of toddy had risen by 60 percent. It can be clearly seen that the per capita consumption of beer had grown at a slow rate of 5 percent between 2000 and 2007.

Per adult consumption of beer, wine, spirits in Sri Lanka is relatively low by world standards. Sri Lanka is ranked within the bottom 25 lowest alcohol-consuming nations, out of a total of 188 countries (Table A2.1, Appendix A2). In 2005, Sri Lanka’s per capita total alcohol consumption of 0.79 litres (of pure alcohol) with a ranking of 168 among the 188 countries. As shown in the table, in terms of pure alcohol, Sri Lankans on average consumed 0.02 litres of beer and 0.33 litres of spirits and a relatively minute amount of wine. However, Sri Lanka’s alcohol consumption depicts a strange and uncommon pattern when compared to the other countries. As seen in Figure 2, the consumption of alcohol in the world indicates that in recent years around 70 percent consume low alcohol content beverages like beer and 10 percent consume wine, while the consumption pattern of hard liquor or high alcoholic beverages seems to be negligible at around 9 percent. This is exactly opposite to the Sri Lankan drinking patterns. For example, consider Figure 10 which presents the share of per capita alcohol consumption by type of legal alcohol. As can be seen, more than 75 percent of the share of per capita alcohol consumption comes from arrack, a high alcohol content beverage, and less than 20 percent of the share of per capita alcohol consumption comes from beer and locally-made foreign liquor.

Table 9 : Consumption of locally produced legal alcohol (litres, million), Sri Lanka, 2000-2007

illustration not visible in this excerpt

Source: Based on various Administration reports of the Excise Commissioner’s General

Table 10: Per capita consumption of pure alcohol by types of alcohol, Sri Lanka, 2000-2007 (proof litres)*

illustration not visible in this excerpt

Note: * Population based on the age group of 21-75 years

Source: Based on various Administration reports of the Excise Commissioner’s General

Figure 10: Share of per capita alcohol consumption by type of legal alcohol, Sri Lanka, 2000-2007*

illustration not visible in this excerpt

Note: * Population based on the age group of 21-75 years

Source: Based on Excise Commissioner’s Annual Reports, various issues.

Governments all over the world tread a fine line when regulating alcohol production. Most governments earn a significant amount of revenue from taxation and the licensing of the alcohol industry. At the same time, exorbitant taxes strike many populations as unfair, and can be used to justify illegal production with associated black market activities (Saunders & de Burgh, 1998). In Sri Lanka, home-based brewing is illegal. However, there is a lucrative unlicensed underground business in most parts of the island. The most commonly used illegal alcoholic beverage in Sri Lanka is known as ‘ Kasippu ’, which is also known to be the cheapest alcoholic beverage available in the country. These beverages are largely outside of government control and as a result, are not taxed (S. Fernando, 2016). In addition, these illegal alcoholic beverages are often not subject to the same standards of quality and purity as its commercially produced counterparts. These illegal alcohol beverages are not produced within a commercial setting and are therefore not reflected in official statistics. However, there are a few studies which give an estimate of the illegal alcohol market in Sri Lanka. For example, WHO (2004) indicated that 92 percent of total alcohol consumption in Sri Lanka is from illegal sources (such as kasippu). (S. Fernando, 2017) highlighted that it has been estimated that there are over 200,000 illicit brew retailers, compared to the 3,200 licensed wine shops (retail outlets) in the country.

Unlike developed countries, less developed countries such as Sri Lanka do not have published data on the percentages of the types of drinkers, be they heavy, moderate or occasional drinkers, except for the information provided in two sample surveys, namely that of the Household Income and Expenditure Survey (HIES) of the Department of Census and Statistics (DCS), and the Consumer Finance Survey (CFS) of the Central Bank which provides only the expenditures on alcoholic beverages at household level. Table 10 presents the average expenditure for one month on alcoholic beverages per spending unit by income groups and Figure 11 presents the data graphically. It is evident from Table 11 that the highest consumed alcohol is arrack followed closely by kasippu. The higher income groups consume more of arrack than any of the other alcohol. This is shown in Figure 11, among all (except for the lowest) income groups, expenditure on arrack is high. The higher income geoups also spend more, in total, on alcohol. As the highlighted, the lowest income group drinkers spend only on kasippu which has a higher alcohol content.

Table 11 : Average expenditure for one month on alcoholic beverages per spending unit by income groups

illustration not visible in this excerpt

Source: Central Bank of Sri Lanka (2004)

Figure 11: Expenditure share for one month on alcoholic beverages per spending unit by income groups

illustration not visible in this excerpt

Source : Based on Central Bank of Sri Lanka (2004)

4 Poverty

There is a large body of literature on poverty in general and on concepts and measurements of poverty in particular. Poverty has also been one of the most debated concepts in development economics and international trade. The international focus on poverty reduction under Millennium Development Goals (MDG) has drawn much more attention to the definition of poverty and its measurements. Over the last two decades or so, the definition of poverty has been a central issue in policy making towards poverty reduction in developing countries. Poverty definitions and measurements have important implications for poverty reduction policies. A study of India and Peru, drawing on both national data and micro surveys, found that significantly different people were identified as poor in the two countries according to two different measurements of poverty ranging from a narrow definition of poverty in terms of the monetary capability to a broader definition of poverty in terms of the participatory approach adopted by the UNDP . Due to the complexities of definitional problems, there is a growing body of literature on definitions of poverty and measurements, while there exists a voluminous literature on poverty itself for poverty definitions). “Fighting poverty in all its dimensions” has become a core developmental policy issue of national governments as well as global institutions such as the WB and the UN .

Existing empirical evidence and various definitions of poverty demonstrate that it is a “multidimensional social phenomenon”. The general discourse carried out relating to poverty includes the material aspects of life, the inclusion of social, cultural and political aspects when studying poverty (S. Fernando, 2015). It also questions what may be achieved, given the available resources and the prevailing environment, or what is actually achieved? Should the same definitions and measurement methods be applied for comparisons in all the countries? Are there “objective” methods, or are value judgments involved? What is the rationale for defining a poverty line? Should it be absolute as in the MDG and in most developing countries, or relative as in the rich OECD countries? (Ehrenpreis, 2006, p. 10).

To address the above questions it is necessary to understand the concepts and measurements that are involved in poverty. The rest of this section focuses on these issues, before discussing the relationship between alcohol consumption and poverty.

Concepts of poverty and its definitions

As noted above the concept of poverty is complex. As a result the term poverty has become “slippery”. Poverty is different things to different people in a country as well as to different people living in different countries. Therefore, it has been defined in a widevariety of ways ranging from the narrow definition in terms of income or consumption to broader definitions in terms of human development, social exclusion, lack of capability in functioning, economic and social vulnerability, livelihood and sustainability, lack of basic needs and relative deprivation. According to the narrow definition in terms of monetary income, the focus has been on whether individuals or households have enough income to meet their basic needs. On the other hand, broader definitions cover other aspects of poverty beyond the level of income and consumption .

Thus, there is no single, universally accepted definition of “poverty”. However, poverty has been traditionally understood to mean a lack of access to resources, productive assets and income resulting in a state of material deprivation[8]. Recently, the concept of poverty and the discussion of its causal explanations have been broadened[9]. As the consumption/income approach to define poverty has come under increased criticism, it has been suggested that in the analysis of poverty, common property resources and the state-provision of commodities should be taken into account and the concept of poverty should be broadened to include the lack of dignity and autonomy . The inclusion of the latter in the meaning of poverty draws from the insight that being non-poor implies a "freedom from the necessity to perform activities that are regarded as subservient and (the) ability to choose self-fulfilling and rewarding life styles” .

Broader definitions of poverty focus on well-being. People who lack the “capability” to function in society might have a lower well-being (S. Fernando, 2016; Sen, 1984) or be more vulnerable to income and price shocks. Thus, poverty means either lack of command over commodities in general (that is, a severe constriction of the choice set (Watts, 1967) or a specific type of consumption (for example, too little food energy intake) deemed essential to constitute a reasonable standard of living in a society, or lack of "ability" to function in a society (Haughton, 2009).

Though Sen in the first instances (1976, 1981, 1985, 1987) emphasized that income was the only valuable factor for increasing the ability to overcome the issues of poverty, in the later instances, Sen (1999) highlighted, by the term “capabilities framework”, that poverty is lack of certain basic capabilities, such as avoiding hunger and illiteracy, as much as lack of adequate income . The new definition formulated by the World Development Report 2000/2001 highlighted, health, education, vulnerability to risk and empowerment are placed alongside economic indicators in the identification of levels and locations of poverty. According to this method, poverty is better measured in terms of different dimensions, most of which are non-economic dimensions, such as basic education, health care, nutrition, water, sanitation, household amenities, as well as in terms of economic dimensions such as income, consumption, employment and wages.

As shown in Figure 12, has proposed the pyramid of poverty concepts to schematize the range of poverty concepts. The top of the pyramid represents the narrowest definition of poverty (private consumption) while the bottom of the pyramid represents the broadest definition. Referring to this pyramid, Haan and Maxwell (1998, p. 4) liken the World Bank's definition of poverty to the top of the pyramid and UNDP's to the base because of the latter's emphasis on human development (S. Fernando, 2015) . One could say the World Bank has adopted a definition of poverty that is close to the top of the pyramid and that UNDP, particularly through its work on human development (UNDP, 1997), has adopted a definition close to the bottom (S. Fernando, 2017) .

As shown in the triangular in Figure 12, the wider the definition of poverty, the richer and more meaningful it is, but the less practical it becomes to operate, and the more difficult it is to make quantitative comparisons.

Figure 12 : Pyramid of Poverty Concepts [10]

illustration not visible in this excerpt

The pyramid of poverty concept provides an important insight into the trade-off between conceptual adequacy and practicality. It seems that income is a much more manageable concept to operate than the more complex multidimensional definitions of poverty. Income is also the central variable in absolute poverty, affecting most or all of the other factors that go into broader definitions of poverty (Institute of Policy Studies, 2005). Hence, most poverty literature point out that the top of the pyramid represents the narrowest definition which is easier to operate while the bottom of the pyramid represents the broadest definition of poverty but difficulty to operate.

Objective and subjective perspectives

The large number of poverty studies carried out defines a poor person as an individual who lacks substantive resources to reach an acceptable standard of living. Generally, the analysis is also restricted to economic deprivation and misery. Although, most researchers seem to agree on this general definition, there have been a large number of discussions about the best approach to measure poverty. Thus, in recent years there have been several studies concerning the effects of using objective (indirect) and subjective (direct) ways of measuring poverty (S. Fernando, Bandara, & Smith, 2016) .

While subjective measures deals with the standard of living people enjoy, the objective measures focus on people’s access to different kinds of resources. For example, subjective indicators are measures of the shortages in the consumption of necessities and of perceived over-indebtedness and scarcity. This is understanding what poverty means and acknowledging that poverty must be defined by those experiencing it themselves; therefore, definitions of poverty may be disempowering and may not capture the true nature of deprivation. On the other hand, objective poverty measures are defined in a more “scientific” manner and focus on quantifiable dimensions; for instance food baskets which meet the nutritional requirements. The main problem of the objective approach is to find a valid and reliable measure of the economic resources people are in control of and defining the boundaries of the poverty line[11].

Chronic and transient poverty

The chronic poor always live in poverty and have very few assets or opportunities to escape it. Groups which are in transient poverty remain so, for short-term periods. The transient poor can move out of poverty once the exogenous shock has passed. For example, transient poverty might be related to seasonality, or to losing a job. The concepts of chronic and transient poverty recognise that poverty is dynamic, and that people may move in and out of it over time.

There are other concepts that come under chronic and transient poverty; namely being chronically poor or transitory poor. If an individual is consistently poor over a long period of time, that person is identified as chronically poor. The transitory poor are identified as individuals who fluctuate between inconsistent periods of poverty. When an individual is in a state of poverty for a specific period, such individuals can be identified as the occasional poor. The non-poor are individuals who consistently live in conditions which are identified to be above the poverty line. These poverty dynamics are shown in Figure 13.

Figure 13: Poverty dynamics

illustration not visible in this excerpt

All in all, poverty is rather complex and multi-dimensional. It is not just an inadequacy of income to meet basic needs or the inability to spend. It also depends on the availability of resources for households or individuals to meet their needs. This aspect is based on the comparison of an individual's income, consumption, education or other attributes which are considered as key factors that contribute to being poor. Poverty is a deprivation of essential assets and opportunities to which every human being is entitled.

Measurements of poverty

Measuring poverty is even more complex than defining poverty and identifying different aspects of poverty. As Maxwell (1999, p. 1) has stated convincingly, “the complexity of measurements mirrors the complexity of definitions, and the complexity increases when participatory methods are used and people define their own indicators of poverty”. Therefore, a large number of poverty measurements have been used in empirical studies since poverty itself is a complex multi-dimensional phenomenon. There are some well-known surveys and reviews on quantitative techniques used in measuring poverty in the poverty literature (Alkire & Foster, 2007; S. Fernando et al., 2016; Lipton & Ravallion, 1993; Sen, 1987, 1999).

There are two common broader concepts of poverty that have been used to measure poverty by researchers and policy analysts. They are absolute poverty and relative poverty. Absolute poverty measures the number of people living below a certain income threshold or the number of households unable to afford certain basic goods and services. If we take one household and if it is below the poverty level income (or expenditure) then that household is in absolute poverty. Relative poverty measures the extent to which a household's financial resources falls below an average income threshold of the economy. Relative poverty compares the lowest segment of a population with the upper segment; it is usually measured by income quintiles or deciles. Trends of these two concepts can move in opposite directions. Absolute poverty can increase while relative poverty decreases, as a result of the gap between the upper and lower strata of a population reducing due to the decline in the well-being of the former which simultaneously makes such households fall beneath the absolute poverty line (World Bank, 2013b, 2013c).

Even within so-called absolute poverty, countries often distinguish between indigence, or primary poverty and secondary poverty (sometimes referred to as extreme and overall poverty). Indigence usually refers to those who do not have access to basic necessities for human survival, while other forms of poverty refer to a degree of deprivation above that threshold. Poverty can be an absolute notion in the space of capabilities, though relative in commodities or characteristics (G. S. Fernando, 2008; S. Fernando, 2015; S. Fernando, Bandara, & Smith, 2013; Sen, 1983). For example, a household incapable of obtaining sufficient food for survival is considered absolutely poor. However, the cost and composition of a food basket may vary considerably between households across different groups, regions and countries. Absolute poverty refers to the lack of ability to satisfy the needs for physical subsistence; and relative poverty extends the concept of poverty to consider individuals as social beings, who have psychological needs to participate in a society and share in its customs and norms.

From all the different aspects and definitions of poverty, analysts have used a large number of empirical measurements or indicators. These include the poverty headcount index, the depth of poverty index, the poverty severity of poverty index, the Foster-Greer-Thobecke poverty index, the human poverty index and the multi-dimensional poverty index.

For a long time, the common practice in measuring poverty has been to count the number of people who fall below some poverty line (defined with varying degrees of arbitrariness) for income or consumer spending. Measuring poverty levels based on the household expenditure is assumed to be more reliable and more stable than using household income (G. S. Fernando, 2008; S. Fernando, 2015; S. Fernando et al., 2013; Klasen, 1997). The poverty literature has highlighted several reasons why the use of income is problematic (Christiaensen, Scott, & Wodon, 2002; Haughton, 2009). For instance, people might forget what they may have earned over the past period and people tend not to disclose the full extent of their income for several reasons. In addition, some sources of income are difficult to observe, such as the extent to which the value of farm output is rising or falling over time. Due to these disadvantages, this study will use the expenditure based poverty line.

Based on the expenditure poverty line, this study will use bivariate and multivariate framework to identify the differences of poverty status among the alcohol consuming household and non-alcohol consuming households. The most widely used poverty indices in the literature belong to the Foster Greer-Thorbecke Index (FGT)[12] to measure the incidence, depth and severity of poverty. FGT measure has been used in capturing the number of poor, depth and severity of poverty among alcohol consuming household and non alcohol consuming households. The FGT Index is defined as:

illustration not visible in this excerpt


n = total population

z = poverty line

yi = income or expenditure of the ith individual

q = the number of persons with income below the poverty line

α= measure of sensitivity of the index to poverty

As a increases, P becomes increasingly sensitivity to the living standard indicator of the poorest people. If α = 0, the FGT Index reduces to the Headcount Index (HI). When α = 1, the index is the Poverty Gap Index (PGI) and if α = 2, the index reduces to the Squared Poverty Gap Index (SPGI) (S. Fernando, 2016) .

illustration not visible in this excerpt

First, the HI is a measure of the incidence of poverty and the simplest measure of poverty. This HI is the proportion of the population for whom per capita income (or other measures of living standard) is less than the poverty line. Even though HI is simple to construct and easy to understand, it disregards differences in the quality of life of different poor households as it assumes that all poor are in the same situation and it does not take the intensity of poverty into account. Further, this index does not account for changes over time, that is if individuals below the poverty line become poorer or richer, as long as they remain below the poverty line (World Bank, 2005).

Second, the PGI indicates how much of income is needed to be transferred to poor individuals in order to allow them to reach the poverty line (S. Fernando et al., 2013). That means this indicator measures the magnitude of poverty, considering both the number of poor people and how poor they are. In addition, the PGI is defined as the ratio of the PG to the poverty line (also called the Foster-Greer-Thorbecke (P1)). It is the poverty gap expressed as a percentage of that line. However, the PG and the PGI have the same deficiencies as the HI; they do not capture differences in the severity of the poverty amongst the poor and ignore “inequality among the poor”.

Third, the SPGI is an indicator which is used to measure the severity of poverty. This index takes inequality among the poor into account . This means that a transfer of any measure of the standard of living from poor to even poorer would reduce the index or a transfer of the same from very poor to less poor would increase the index. Therefore, it is difficult to interpret the poverty gap of the poor with the index itself. Further, it is the average value of the square of the depth of poverty for each individual. Poorest people contribute relatively more to the index (also called the Foster-Greer-Thorbecke (P2)). The poverty severity index provides a weight to the poverty gap. Larger values of the parameter indicate that a greater weight is attached to the poverty gap of the poorest unit. The SPGI is defined as the average of the square relative to the poverty gap of the poor.

Given the various sources of arbitrariness in poverty measurement, one can still draw upon stochastic dominance criteria to make unambiguous comparative rankings of poverty. One of the most commonly used measurements of a welfare improvement indicator is the Gini Coefficient (GC) and usually measured as follows:

illustration not visible in this excerpt

The value of the Gini Coefficient is bound between 0 and 1. The value 0 indicates the case of full equality and one indicates the case where there is complete inequality and all income (or expenditure) accrues to a single individual. Hence, low values of the Gini are associated with more equal distribution of income or expenditure.

In addition to the above mentioned traditional poverty measures, new qualitative as opposed to quantitative, approaches to poverty assessment have emphasized the poor’s own criteria of poverty as well as their own solutions (Chambers, 1997). The policy implications of this new approach have emphasized programmes that enable the poor to exercise their agency. Their solutions are creating a viable environment by making available critical external resources such as credit. Empowerment of the poor is viewed as critical to the success of poverty elimination (S. Fernando, Bandara, Smith, & Pham, 2015). Another recent approach, the social exclusion approach, emphasizes the importance of institutions and norms that exclude certain groups from a variety of social networks and the importance of social solidarity in sustaining livelihoods[13]. These new insights on poverty have far-reaching implications for analyzing the gendered nature of poverty as well as the relationship between gender inequalities and overall poverty levels (Haan & Maxwell, 1998; ILO & UNDP, 1996; UN, 1994).

More recently, the United Nations Development Programme Human Development Report introduces the Multidimensional Poverty Index (MPI). The MPI reveals a different pattern of poverty than income poverty, as it illuminates a different set of deprivations. As shown in Figure 14, MPI has three dimensions and 10 indicators (S. Fernando, 2015, 2016, 2017). Each of the three dimensions constitute one-third of the total and each indicator within each dimension is equally weighted (Alkire & Foster, 2007). A person is considered poor if they are deprived in at least 30 percent of the weighted indicators. The intensity of poverty denotes the proportion of indicators in which they are deprived.

Figure 14: Dimensions and Indicators of the MPI

illustration not visible in this excerpt

Source: UNDP (2010).

The MPI can be broken down to show a vivid picture of people living in poverty, both across countries, regions and the world and within countries by ethnic group, urban/rural location, or other key household characteristics. It is the first international measure of its kind, and offers a valuable complement to income poverty measures because it measures deprivations directly. The MPI can be used as an analytical tool to identify the most vulnerable people, to show aspects in which they are deprived and help to reveal the interconnections among deprivations (OPHDI, 2010).

However, similar to other measurements the use of this index is still debatable. Ravallion (2010, p. 1) points out that there are highly questionable value judgments in the MPI. For example, how can one contend (as the MPI does implicitly) that the death of a child is equivalent to having a dirt floor, cooking with wood, and not having a radio, television, telephone, bike or car? Or that attaining these material conditions is equivalent to an extra year of schooling (such that someone has at least 5 years of schooling) or to not having any malnourished family member?

Nonetheless, the MPI attempts to capture multidimensional aspects of poverty. It also attempts to measure the real chronic poor, poor and non-poor of any country by adopting not only income or expenditure but also incorporate the many dimensions related to poverty such as health, social and economic dimensions of poverty.

4 Global Poverty

The discussion in the previous sections indicates that poverty is a complex and multi-dimensional phenomenon. Measuring poverty is even more complex and controversial. Absolute poverty (head count index), however, has now become a commonly used and widely accepted measurement in assessing overall performance of poverty reduction at the global level, particularly in developing countries (Chen & Ravallion, 2007, p. 2). There are different estimates of absolute poverty in the literature because of complexities involved in poverty and measurement problems (Kaplinsky, 2005). Although data on absolute poverty has often been questionable, there are two methods of measuring absolute poverty, either by using national accounts data or by using national survey data. In this section, internally consistent time-series data on absolute poverty for ten reference years developed by the World Bank (Chen & Ravallion, 2007) are used to provide an over view of global trends in poverty over the last three decades or so.

The World Bank defines absolute poverty based on the purchasing power parity (PPP) exchange rate. The World Bank defines extreme poverty using a ‘US$1.25-a-day’ poverty line and moderate poverty using a ‘US$2.00 a day poverty line’. Table 12 provides both the proportion and number of people living below $1.25 and $2.00 a day for different regions of the world from 1981 to 2008. As shown in Table 12, there has been a rapid decline in the absolute poverty rate in the East Asia and Pacific region. Figure 15 further demonstrates that, in this region, while the proportion of the people living below $1.25 has fallen from 77 percent to 14 percent, the proportion of the people living below $2.00 has fallen from 92 percent to 33 percent in this period. The figures, in Table 12, clearly demonstrate that the East Asia and Pacific, Sub-Saharan Africa and South Asia regions have managed to reduce absolute poverty significantly during the last three decades. For example, in South Asia extreme poverty decreased from 61 percent in 1981 to 36 percent in 2008. Figure 16 shows the absolute numbers of the poor population in each continent. Although South Asia have managed to reduce absolute poverty significantly, in 2008, the region had the highest number of extreme poor population (around 570 million). The extreme poor and moderate poor population in South Asia has increased form 810 million in 1981 to 1,120 million in 2008; indicating that South Asia is home to the largest number of poor in the world. Similar to South Asia, the absolute numbers of the extreme and moderate poor population in the Sub-Saharan Africa region have increased from 204 million in 1981 to 386 million in 2008.

Figure 17 shows the share of the regional population below the $1.25 and $2.00 poverty lines. It is clear that the share of poor people in South Asia and Sub-Saharan Africa has increased compared with the other continents. South Asia and Sub-Saharan African continents are home to a very large proportion of the poor in the world. In 1981, 30 percent of the share of poor people was from the South Asia continent and this share increased to 45 percent in 2008. In terms of the Sub-Saharan Africa, in 1981, 11 percent of the share of poor people resided in that continent and this has more than doubled in 2008. Therefore, although the aggregate poverty rate in these regions has fallen over the last three decades, the proportion of poor people has increased in the South Asia and Sub-Saharan Africa continents.

Table 12: Poverty across the globe

illustration not visible in this excerpt

Source: Table extracted form (http://data.worldbank.org/topic/poverty) on August 05, 2013.

Figure 15: Poverty Headcount Ratio. Country groups, 1981-2008

illustration not visible in this excerpt

Source: Based on the data extracted form (http://data.worldbank.org/topic/poverty) on August 05, 2013.

Figure 16: Poor Population. Country groups, 1981-2008

illustration not visible in this excerpt

Source: Based on the data extracted form ( http://data.worldbank.org/topic/poverty) on August 05, 2013.

Figure 17: Share of the regional population below $1.25 and $2.00 poverty lines

illustration not visible in this excerpt

Source: Figure extracted form (http://data.worldbank.org/topic/poverty ) on August , 2013.

As seen from Table 12 and Figure 17, currently more than 95 percent of the poor in the world are living in three regions: East Asia and Pacific, South Asia, and Sub-Saharan Africa. Even if the poverty line is defined as $2.00 a day, these three regions would still account for more than 90 percent of the poor for the 1981 to 2008 period. In the East Asia and South Asian regions, the bulk of those in poverty are from China and India.

Table 13 presents, for 5 South Asian countries, the percentage of people living below the national and international poverty lines, as measured by the Gini coefficient and the country’s population (in millions). As can be seen from column (8), the South Asian region has a population of over 1.5 billion and India is the most populous country and Sri Lanka is the least populated among the 5 countries. A comparison of the percentage of population living in poverty, from columns (5) and (6) based on international poverty lines, demonstrates that Bangladesh and Nepal?? have the highest rate of poverty and Sri Lanka has the lowest rate of poverty. Based on the poverty lines of the individual countries, the poverty rate is 40 percent in Bangladesh, 30.9 percent in Nepal, 27.5 percent in India, 22.3 percent in Pakistan and 15.2 percent in Sri Lanka. According to the World Bank (2012) report, 29.1 percent of Sri Lanka’s population lives on under $2.00 a day in 2007 compared with 7 percent living under $1.00 a day. Among the 5 countries, Sri Lanka has the least people living under poverty. However, comparing the Gini coefficients, income inequality is highest in Sri Lanka than among the other four countries in the region.

Table 13: Poverty rates in the South Asia region

illustration not visible in this excerpt

Source: World Bank (2012)

Figure 18 shows the changes in Gini coefficient in selected South Asian Countries from the 1980s to the 21st century. Except for Pakistan, the Gini coefficients for the other four countries increased until 2002. In particular, the Gini coefficients increased sharply in Sri Lanka and Nepal during the 1984 to 2002 period. However, after the mid-2000s most of the South Asian countries’ Gini coefficients decreased except for India. Income inequality in Sri Lanka rose significantly from 0.32 in 1991 to 0.41 in 2002. Although the Gini coefficient in Sri Lanka decreased slightly to 0.36 in 2010, at present Sri Lanka has been recorded as having the highest income inequality compared to the other countries in the region.

Figure 18: Changes in the Gini coefficient in selected South Asian countries

illustration not visible in this excerpt

Source: Numbers extracted form ( http://data.worldbank.org/topic/poverty) on August 08, 2013.

5 Poverty in Sri Lanka

In the previous section, the overall trends in global poverty were highlighted. This section focuses on poverty in Sri Lanka, in detail, in order to lay the foundation for this study. Sri Lanka is a developing country in South Asia with better living conditions and relatively lower poverty levels than its neighbours (see Table 13). Following the common and traditional practice, used over decades in policy analysis, the location of households in Sri Lanka can be categorized into three regions, Urban, Rural and Estate. Areas governed by either the municipal council or urban council is considered as the urban region. Plantation states with tree crops, namely; tea, rubber and coconut, spanning more than 20 acres and having not less than 10 residential labourers, are considered as the estate region. Residential areas, which do not belong to the urban or estate regions, are considered as the rural region. The socioeconomic characteristics of households in these three regions are significantly different.

Table 14 demonstrates the basic characteristics of the above three sectors in terms of the distribution of population, monthly income and the level of education.

Figure 14: Global Shares of Alcohol Production, 1961-2009

illustration not visible in this excerpt

Note: Excluding Jaffna, Mannar, Kilinochchi and Mullaithivu districts in the Northern Province.

Source: Based on the DCS (2010)

As shown in Table 14, over 80 percent of Sri Lankans (or about 15.8 million) are still living in the rural sector. While a low percentage of 14.2 percent of the total population (2.8 million) are living in the urban sector, only 5.6 percent of the total population are living in the estate sector (1.1 million). The overall average household size for Sri Lanka is 4 people. However, households in the estate sector tend to have a larger family size compared to the urban and rural sectors. Table 14 further shows that the household monthly income in Sri Lanka is around Rs. 35,495. Comparing the three sectors, the urban sector has the highest average household monthly income (Rs. 46,196) and the estate sector has the lowest average household monthly income (Rs. 25,649). As shown by the education figures presented in Table 14, it is clear that most uneducated household heads live in the estate sector as the estate sector holds the highest percentage of no schooling at 10 percent and only up to grade 5 at 53 percent. In addition, the highest share of food expenditure comes from the estate sector at 49.5 percent; thus, the estate sector spend more on food than the households in urban and rural sectors. All these values confirm that the living conditions for households in the estate sector is the worst of all the three sectors in Sri Lanka.

A number of measurements of poverty are widely available for Sri Lanka. Figure 19 illustrates the trends in poverty since the 1990s for households at the sectoral and national levels, in Sri Lanka (DCS, 2011).

Figure 19: Headcount index by Sector

illustration not visible in this excerpt

Source: Based on data from DCS, various years.

The incidence of poverty (as measured by the standard headcount ratio) at national level has shown a steady decline from 26.1 percent in 1990/1 to 8.9 percent in 2009/10. In other words, poverty has decreased by 66 percent during this period. However, the gap in poverty incidence between the sectors widened from 1990/91 to 2006/07 before decreasing in 2009/10. Urban and rural poverty declined by 59 percent and 47 percent, from 1990 to 2006, respectively. in contrast, the inverse trend is apparent in the estate sector, with poverty increasing by about 56 percent for the same period.

Poverty in Sri Lanka is predominantly a rural phenomenon. Although poverty level in the estate sector is very high, the share of total population in that sector is relatively small (Figure 19). Therefore, poverty analysts very often label Sri Lankan poverty as a rural phenomenon. Several studies have shown that poor households are more likely to be found in the rural than in urban areas due to working members being employed in agriculture and other primary production activities (Datt & Gunewardena, 1997). At the sectoral level, poverty in the rural sector, where 80 percent of the population derives their livelihood, first increased from 29.5 percent in 1990/91 to 30.9 percent in 1995/96 but then continued to fall to 9.4 percent in 2009/10. The sharp drop of rural poverty from 2002 is the main contributor for the unprecedented drop of poverty at the national level. However, the estate sector which accounts for about 5.5 percent of the population saw their poverty increasing from 30 percent in 2002 to 32 percent in 2006/07. Nevertheless, in 2009/10, the estate sector’s poverty rate has decreased by two-third’s as a result of the drop in the relative prices of food items and an increase in employment and wages.

Table 14: Poverty headcount index (percentage) by district

illustration not visible in this excerpt

Source: DCS (2011)

For administrative purposes, Sri Lanka is divided into 22 districts. These are listed in Table 15. Table 15 presents the poverty head count index for the 3 regions and for the 22 districts. Intra-district variations of the poverty head count index from 1990/91 to 2009/10 are depicted in Map 2.1. This map shows that for every district, poverty has decreased considerably. In addition, significant reductions of poverty were recorded in every district, in 2009/10, except for the Batticaloa and Ampara districts. The highest percentage of poor households in 2009/10 was from the Batticaloa (20.3 percent) and Jaffna (16.1 percent) districts; while the lowest percentage is reported from the Colombo district at 3.6 percent. Table 15 reveals that, in 2006/07, the poverty head count ratio for Batticaloa and Ampara was 10.7 and 10.9 percent, respectively. However, in 2009/10, these two districts’ poverty head count ratio rose to 20.3 and 11.8 per cent, respectively. Furthermore, Hambantota district, which had showed a poverty reduction of as much as 60 percent between 2002 and 2006/07, showed a further 46 percent fall in the poverty head count ratio between 2006/07 and 2009/10. Nuwara-eliya, Moneragala, Rathnapura and Badulla, the poorest four districts in 2006/07, reported around 50 percent improvement in 2009/10.

Map 2.1: Poverty Head Count Index by Districts – 1990/91, 1995/96, 2001/02, 2006/07 and 2009/10

illustration not visible in this excerpt

Source: Illustrations based on HIES 1990/91, 1995/96, 2001/02, 2006/07 and 2009/10.

Although poverty has significantly decreased in recent year, studies shows many people who have managed to get out of poverty are still at risk of slipping back, as a large proportion of them are just above the poverty line. For example, if the poverty line is increased by 10 percent, the percentage of poor increases to 12.8 percent (HIES 2009/10), which amounts to an increase of around 800,000 poor people (Nanayakkara, 2012). Therefore, any short or long term economic shock, due to whatever reason, may push them back to poverty.

Income Distribution

Average income per month per spending in Sri Lanka has increased rapidly from Rs. 881 in 1980/81 to Rs. 35495 in 2009. During this period, income of all the three sectors has increased considerably. For instance, during this period, urban income increased from Rs. 1274 to Rs. 46196, rural income increased from Rs. 795 to Rs. 34329 and estate income has risen from Rs. 753 to Rs. 25649. Though all the income groups enjoyed an increase in income, increases in income was not evenly distributed among all the income groups.

Table 15: Income Distribution , 1980-2009, selected years

illustration not visible in this excerpt

Note: n.a denotes not available.

Source: DCS (2011), various years.

As shown in Table 16, since 1990, the share of income of the poorest 40 percent of the spending units increased form 14.8 percent to 15.2 percent in 1995/96 and then progressively fell to 13.2 percent in 2006/07. During this period, the share of income of spending units in the richest 20 percent declined from 51.4 percent in 1990/91 to 50.3 percent in 1995/96 and then continuously increased to 54.7 per cent in 2006/07. Thus, the richest 20 percent has received more than 50 percent of the total household income in all three surveys over the period while the share of the poorest 40 percent has remained around 14 percent. However, average income per spending unit shows that all the income groups have benefited from an increase in income even though the increases were not evenly distributed across the income groups. The distribution of income suggests that most of the increased income has accumulated in the hands of the upper income deciles.

Figure 20: Gini Coefficient index by sector

illustration not visible in this excerpt

Source: Based on data from DCS, various years.

Since Sri Lanka’s independence in 1948, income distribution in Sri Lanka, as measured by the Gini coefficient, has moved towards lower inequality. However, since 1980, Sri Lanka’s income inequality has worsened with the Gini coefficient for spending units increasing from 0.43 in 1980/81 to 0.47 in 2009, and reaching a high of 0.49 in 2006/07 (Figure 20). The Gini coefficient for the sectors shows, too, that the distribution of income is also uneven within sectors. The urban sector displayed the highest inequality in 1990/91, reaching a high of 0.62, but falling again, in 2009, to almost the same coefficient in 1980/81. The lowest inequality, in the earlier years, can be seen in the estate sector where the Gini coefficient has only fluctuated between 0.25 and 0.34. However, in 2006/07, the coefficient peaked at 0.57 to exceed that for the other two sectors. It then fell, in 2009, to approximately that for the other two sectors. In other words, the estate sector has had the highest growth in income inequality. The Gini coefficient has shown the least fluctuations for the rural sector. For the period, 1980 to 2009, the data clearly shows that the disparities between the rich and the poor have remained stagnant, for Sri Lanka as a whole, throughout the period from 1980/81 to 2009.

6 Relationship between Alcohol Consumption and Poverty

The harmful effects of excessive alcohol consumption in low- and middle-income countries have created substantial economic cost to individuals as well as to the country. The consumption of alcohol reduces income available for the essentials of life, such as food, and is thus a major contributor to poverty and poor health (WHO, 2008).

In addition to the above, the excessive consumption of alcohol has also contributed to the incidence of poverty. Excessive alcohol consumption has both a direct and an indirect effect on the poverty status of individuals, families, and the entire community. Excessive consumption has negative impacts on the achievement of key human development outcomes in health and education, and is a key driver for poverty in the long-run. In Sri Lanka, alcohol consumption is found to be higher among poor families compared to non-poor families (Jayathilaka, 2007). Facts on household alcohol consumption expenditure are not extensive. However, there is evidence that alcohol consumption crowds out essential items in low-income households (Pu, Lan, Chou, & Lan, 2008). In Britain, the 2006 Family Expenditure Survey reported an average expenditure on alcohol consumption for all households of £14.8 per week (Institute of Alcohol Studies, 2007).

Figure 20 provides an overview on various ways in which alcohol consumption may affect poverty. Poverty may be primarily determined by labour market outcomes, but it is also affected by family composition. Family composition affects poverty by altering family size, and sources and quantity of non-earned income. Labour market outcomes are determined by a person's human capital, which in this case is summarized by a person's level of education and other human capital investments (such as training and health). For example, a single parent may not be able to work as a childless person (Korenman & Neumark, 1991; Neumark & Korenman, 1992). Alcohol consumption and poverty are related because alcohol consumption affects the determinants of poverty, namely that of education, human capital investments, marriage, and family composition. Finally, person specific factors such as ability, preferences, and family background affect alcohol consumption, as well as educational achievement, skill accumulation, marriage and health.

Figure 21: A simple behavioural model of alcohol consumption and poverty

illustration not visible in this excerpt

For the most part, the implied relationships in Figure 2 are obvious and consistent with intuition; the prime example of this statement being the effect of alcohol consumption on human capital. The physiological effects of alcohol consumption, particularly those related to chronic alcohol consumption, suggest that alcohol consumption is expected to result in a reduction of physical and cognitive abilities. Consequently, alcohol consumption is expected to lower productivity, reduce earnings, and result in an increased likelihood of poverty. Similarly, alcohol consumption may adversely affect educational achievement, or attainment, and hence lower earnings and increase poverty. In addition, it is obvious that alcohol consumption affects poverty through its effect on health and other means of livelihood.

Figure 21 also illustrates that an individual’s ability, preferences, and family background determines both poverty and alcohol consumption. For example, a person with a high rate of time preference is more likely to use alcohol, make fewer human capital investments, and experience more poverty than would a similar person with a lower rate of time preference. Thus, it is important to control these factors if the objective is to estimate a causal effect of alcohol consumption on poverty.

However, there are two issues that Figure 2 ignores. The first is that poverty may cause alcohol consumption. This possibility is most relevant if poverty is primarily a demand-determined phenomenon where opportunities for work and pay are limited[14]. In these circumstances, alcohol consumption may be encouraged by the absence of significant positive returns on human capital investments. Alcohol consumption may adversely affect human capital development and, as a result, economic development. Therefore, in addition to the direct monetary cost of alcohol consumption, there is another cost of alcohol consumption that is associated with a diminished level of human capital and lower earnings. In areas where there is limited economic opportunity and relatively low returns on human capital investment, the full price of alcohol consumption is low, and as a result, alcohol consumption is more likely to occur.

In addition, alcohol has been identified as one of the most harmful products to human health. Studies have found that around 4 percent of the burden of diseases worldwide is due to alcohol consumption (Rodgers et al., 2004; Room, Babor, & Rehm, 2005). In addition, alcohol consumption is estimated to contribute to more than 60 health related problems. These include chronic diseases such as cirrhosis of the liver and certain cancers, as well as poor health resulting from trauma, violence, and accidental injuries. Therefore, most governments try to control the consumption of alcohol through laws, although very few countries ban alcohol consumption entirely. On the other hand, the World Health Organisation (WHO) has recommended that its member countries should develop national control policies to prevent excessive alcohol consumption. Nonetheless, there are significant variations to be found between national alcohol policies.

Many studies in Sri Lanka have documented that alcohol consumption is an important factor in the persistence of poverty in the country (Room & WHO, 2002). A large number of families are found to be unable to escape from poverty because of the alcohol consumption, and tobacco use, of one or more of their members. Once influenced by alcohol consumption, poverty itself may be exacerbated by absenteeism, lack of motivation, ill-health and lack of family unity (Baklien & Samarasinghe, 2005). It is a leading risk contributor to the burden of disease and a key cause of disability, especially among men. Thus, excessive alcohol consumption generates a wide range of interrelated negative effects and outcomes, some primary and others secondary. Most common primary consequences are domestic violence, violence against women, injuries and accidents, petty crime, violence, and declines in labour productivity.

From the health activists' point of view, alcohol consumption has resulted in serious health and safety issues that adversely affect society. The WHO highlights that the effect of alcohol consumption on the health of the population is the third biggest issue in Sri Lanka, followed by premature death caused by tobacco, and high blood pressure (WHO, 2004). While there is no guarantee that abstaining from alcohol consumption will reduce poverty levels it will certainly help to reduce the health hazards, violence, and indebtedness of individuals and households – indirectly helping to sustain income levels of affected households.

7 Concluding Remarks

The purpose of this chapter was to provide the background for the thesis in terms of alcohol consumption and poverty. This chapter provides an overview of the different types of alcoholic beverages. Following the general description of alcohol, the global pattern of alcohol consumption was examined. It has shown that although global production of the alcohol beverages has grown rapidly, the share of fermented beverages and wine production has decreased in the recent years. In terms of the pure alcohol content, beer and wine consumption was dominated by the European continent and spirits consumption was dominated by the North American continent. Consumption of beverages other than wine, beer and spirits were dominated by the African countries. A comparison of each type of alcohol beverages consumption in LDC demonstrated that consumption of beer and spirits has increased at a higher rate compared to wine. In DC, there has been an increasing rate of wine consumption compared to consumption of other alcoholic beverages.

Sri Lanka’s alcohol production and consumption have grown rapidly and more than 75 percent of the government’s revenue, from alcohol taxes and levies, comes from arrack production. As well, hard liquor consumption is relatively high in Sri Lanka compared with the global trend. Illegal alcohol (mainly known as Kasippu) consumption in Sri Lanka is popular among people because it is cheaper.

In the second half of this chapter, various definitions on poverty and its measurements were briefly discussed in order to emphasise that the term ‘poverty’ is not merely a lack of income, and to broaden the term’s connotations. This chapter also served to highlight the fact that poverty is a complex phenomenon with various social, health and economical influences. Following a survey of definitions of poverty and description of poverty measurements, this chapter provided an overview of global trends in poverty and poverty in Sri Lanka. Although the aggregate poverty rates declined in the world, the number of poor people has increased in South Asia and Sub-Saharan Africa continents. Compared to the other South Asian countries Sri Lanka has had better living conditions. There is a declining trend in poverty in Sri Lanka. However, the income inequality has grown in Sri Lanka in recent years as it has been in the South Asian region.

Consumption of alcohol may affect the determinants of poverty through the education, health, human capital and other various socio-economic and demographic characteristics. In the next chapter, this study will review the biological, psychological, and socio-cultural factors that impinges on alcohol consumption and poverty.


Table A 1: Countries with the highest per adult consumption of total alcohol, beer, wine, spirits in pure alcohol (in litres), 2005

illustration not visible in this excerpt

Source: Based on WHO (2011)

Table A2 Liquor Manufacturing Institutions and Companies - 2012

illustration not visible in this excerpt

Table A2: Liquor Manufacturing Institutions and Companies – 2012 (Contd…)

illustration not visible in this excerpt

Note: 1 Enclose malt liquor beer which has 5 per cent below and above alcohol per cents.

2 Includes molassess arrack, coconut and processed arrack and special arrack.

3 Comprise distilled and manufactured spirits from coconut toddy, palmyrah toddy and rectified spirits.

4 Contains only Country made foreign liquor.

Source: Excise Department of Sri Lanka (2012)


Alkire, S., & Foster, J. (2007). Counting and Multidimensional Poverty Measurement. Retrieved from

Baklien, B., & Samarasinghe, D. (2005). Alcohol and poverty in Sri Lanka. Colombo: FORUT.

Baulch, B. (1996). Editorial: The New Poverty Agenda: A Disputed Consensus1. IDS Bulletin, 27 (1), 1-10. doi:10.1111/j.1759-5436.1996.mp27001001.x

Central Bank of Sri Lanka. (2004). Consumer Finance and Socioeconomic Survey 2003/2004. Retrieved from

Chambers, R. (1997). Whose Reality Counts?: Putting the First Last London: Intermediate Technology.

DCS. (2011). Poverty Indicators - Household Income and Expenditure Survey - 2009/10. Retrieved from Department of Census and Statistics, Sri Lanka:

Ehrenpreis, D. (2006). What is poverty? Concepts and measures. Retrieved from http://econpapers.repec.org/RePEc:ipc:ifocus:9

Excise Department of Sri Lanka. (2012). Administrative Report of the Commissioner General of Excise for the Year 2012. Retrieved from No. 34, W.A.D. Ramanayeka Mawatha, Colombo 02, Sri Lanka.:

Fernando, G. S. (2008). An Assessment of the Influence of Fundamental Factors on Share Prices in Sri Lanka.

Fernando, S. (2015). Tourism in Sri Lanka and a computable general equilibrium (CGE) analysis of the effects of post-war tourism boom. (Unpublished Ph.D Thesis), Griffith University, Griffith Bussines School. Retrieved from https://experts.griffith.edu.au/publication/n6e5a04ffcf492b955bd3062eca4db2cb

Fernando, S. (2016). Managing the Post-War Tourism Development in Sri Lanka. International Journal of Business and Social Science, 7 (11).

Fernando, S. (2017). The Tourism-Led Development Strategy in Sri Lanka. Journal of Business and Thechnology, 1 (1).

Fernando, S., Bandara, J., & Smith, C. (2013). Regaining missed opportunities: the role of tourism in post-war development in Sri Lanka. Asia Pacific Journal of Tourism Research, 18 (7), 685-711. doi:10.1080/10941665.2012.695284

Fernando, S., Bandara, J., & Smith, C. (2016). Tourism in Sri Lanka. In M. C. Hall & S. J. Page (Eds.), The Routledge Handbook of Tourism in Asia (pp. 251-264). Abingdon,Oxon, UK: Routledge.

Fernando, S., Bandara, J., Smith, C., & Pham, T. (2015). SLCGE-Tourism: a computable general equilibrium model of the Sri Lankan economy for tourism policy analysis. Griffith Business School, Discussion Papers - Economics, 2015-06, 33. Retrieved from Griffith University website: https://www120.secure.griffith.edu.au/research/items/694206df-16db-4622-b239-c1b60cae4de8/1/

Foster, J. E., Greer, J., & Thorbecke, E. (1984). Notes and Comments A Class of Decomposable Poverty Measures. Econometrica, 52 (3), 761.

Haan, A. d., & Maxwell, S. (1998). Editorial: Poverty and Social Exclusion in North and South. IDS Bulletin, 29 (1), 1-9. doi:10.1111/j.1759-5436.1998.mp29001001.x

Haughton, J. H. (2009). Handbook on poverty and inequality / Jonathan Haughton, Shahidur R. Khandker. Washington, DC :: World Bank.

ILO, & UNDP. (1996). Social Exclusion and Anti-Poverty Strategies. Retrieved from

Institute of Alcohol Studies. (2007). Alcohol: price, legal availability and expenditure: IAS Factsheet. Retrieved from http://www.ias.org.uk/resources/factsheets/price_availability.pdf

Kangas, O., & Ritakallio, V.-M. (1998). Different Methods – Different Results? Approaches to Multidimensional Poverty, Empirical poverty research in a comparative perspective / edited by Hans-Jurgen Andress. Aldershot, Hants, England ; Brookfield, Vt. : Ashgate.

Kaplinsky, R. (2005). Globalization, poverty and inequality: between a rock and a hard place. Malden, MA: Polity.

Klasen, S. (1997). Poverty, Inequality and Deprivation in South Africa: An Analysis of the 1993 Saldru Survey. Social Indicators Research, 41 (1/3), 51-94. doi:10.2307/27522257

Korenman, S., & Neumark, D. (1991). Does Marriage Really Make Men More Productive? Journal of Human Resources, 26 (2), 282-307.

Lipton, M. (1997). Editorial: Poverty -- are there holes in the consensus? World Development, 25 (7), 1003-1007. doi:Doi: 10.1016/s0305-750x(97)00031-4

Lipton, M., & Ravallion, M. (1993). Poverty and policy. Washington, D.C: Policy Research Department, World Bank.

Maxwell, S. (1999). The Meaning and Measurement of Poverty,. Retrieved from Overseas Development Institute, Portland House, Stag Place, London SW1E 5DP: http://www.odi.org.uk/sites/odi.org.uk/files/odi-assets/publications-opinion-files/3095.pdf

Nanayakkara, W. (2012). Eradicating Poverty in Sri Lanka: Strong Progress but Much Remains To Be Done. Retrieved from Colombo 7, Sri Lanka:

Neumark, D., & Korenman, S. (1992). Sources of Bias in Women's Wage Equations: Results Using Sibling Data. Retrieved from http://ideas.repec.org/p/nbr/nberwo/4019.html

OPHDI. (2010). Multidimensional Poverty Index. Retrieved from http://www.ophi.org.uk/policy/multidimensional-poverty-index/

Penttilä, I., & Nordberg, L. (n.d.). Objective and Subjective Measures of Poverty in the European Community Household Panel. Retrieved from http://www.stat.fi/isi99/proceedings/arkisto/varasto/nord0702.pdf

Pu, C., Lan, V., Chou, Y., & Lan, C. (2008). The crowding-out effects of tobacco and alcohol where expenditure shares are low: Analyzing expenditure data for Taiwan. Social Science & Medicine, 66 (9), 1979-1989. doi:DOI: 10.1016/j.socscimed.2008.01.007

Ravallion, M. (2010). Guest Blog: World Bank research director critiques the new UN poverty index. Retrieved from http://www.oxfamblogs.org/fp2p/?p=3070

Rodgers, A., Ezzati, M., Vander Hoorn, S., Lopez, A. D., Lin, R.-B., Murray, C. J. L., & Group Comparative Risk Assessment, C. (2004). Distribution of Major Health Risks: Findings from the Global Burden of Disease Study. PLoS Med, 1 (1), e27.

Room, R., Babor, T., & Rehm, J. (2005). Alcohol and public health. The Lancet, 365 (9458), 519-530.

Saunders, J. B., & de Burgh, S. (1998). The distribution of alcohol consumption. In M. Grant & J. Lifvak (Eds.), Drinking Patterns and their Consequences. Washington: DC: Taylor and Francis.

Sen, A. (1976). Poverty: An Ordinal Approach to Measurement. Econometrica, 44 (2), 219-231.

Sen, A. (1981). Public Action and the Quality of Life in Developing Countries. Oxford Bulletin of Economics & Statistics, 43 (4), 287-319.

Sen, A. (1999). Development as freedom New York :: Knopf.

UN. (1994). Improving Concepts and Methods for Statistics and Indicators on the Situation of Women. Retrieved from New York:

UNDP. (1997). Human Development Report-1997. Retrieved from http://hdr.undp.org/en/reports/global/hdr1997/chapters/

WHO. (2008). Investing in Mental Health. Retrieved from World Health Organization, Department of Mental Health and Substance Dependence, Avenue Appia 20, 1211 Geneva 27, Switzerland: http://www.who.int/mental_health/media/investing_mnh.pdf

WHO. (2011). Global Status Report on Alcohol and Health - 2011. Retrieved from World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland: http://www.who.int/entity/substance_abuse/publications/global_alcohol_report/msbgsruprofiles.pdf http://www.who.int/substance_abuse/publications/global_alcohol_report/profiles/lka.pdf

World Bank. (2005). Poverty Manual, Poverty Indices: Comparison of Measures. Washington: World Bank Group, World Bank.

World Bank. (2012). World Development Report 2012: Gender Equality and Development: The International Bank of Reconstruction and Development/ The World Bank, 1818 H Street NW, Washington, DC 20433.

World Bank. (2013a). Data Catalog. Retrieved from: http://datacatalog.worldbank.org/

World Bank. (2013b). Poverty. Retrieved from http://www.worldbank.org/en/topic/poverty


[1] Asia Pacific Research Institute, Research Fellow.

[2] This is the most comprehensive summary available for world alcoholic production trends; although it may be possible to obtain partial information of individual beverages for some earlier years. Also, since it is impossible to obtain reliable data for the production of traditional fruits or cereal-based alcoholic beverages in Africa, Asia and Latin America, the data are limited to the more commercialized alcoholic beverages.

[3] Another possible reason for the fall in the wine shares and increase in that of beer could be due to land area and weather limitations associated with wine production compared to the mostly technology-based beer production.

[4] In our analysis, we have excluded countries, especially countries that have laws which ban the consumption of alcohol for religious reasons.

[5] The definition of LDCs is from the United Nations list of countries. Least developed countries are also included in the Less developed countries (LDC).


[7] Government revenue represents revenue collected from excise duties on manufactured alcohol.

[8] See (Baulch, 1996) for more discussion of the definition of poverty.

[9] This discussion draws heavily from (Baulch, 1996; Lipton, 1997)

[10] Based on (USAID, 2004)

[11] This discussion draws heavily from (Kangas & Ritakallio, 1998; Penttilä & Nordberg, n.d.)

[12] Based on the original study done by (Foster, Greer, & Thorbecke, 1984)

[13] The social exclusion approach was first developed in France to address poverty in the context of Western Europe. It is increasingly being used for understanding poverty in developing country contexts as well.

[14] This previous discussion ignores the effect of income on alcohol consumption. Depending on whether alcohol consumption is a normal or inferior good, income will either be positively or negatively correlated with alcohol. In either case, the direction of causality is from alcohol consumption to poverty.


ISBN (Book)
File size
1.2 MB
Catalog Number




Title: Alcohol consumption and poverty in Sri Lanka