Information about Europe Europa
Journal Issue
Share
Article

Albania: Poverty Reduction Strategy Paper—Annual Progress Report

Author(s):
International Monetary Fund
Published Date:
January 2006
Share
  • ShareShare
Information about Europe Europa
Show Summary Details

3 Socioeconomic conditions

This section uses statistical data to look at some key aspects of socio-economic conditions in 2002 and 2003. The main point of interest is to update the poverty profile. Evidence on education, health, the labour market and utilities is also presented. The main source of information is the Living Standards Measurement Survey (LSMS) carried out by INSTAT.

3.1 POVERTY

Last year’s Progress Report presented the evidence on the level of poverty in Albania using the data of the 2002 LSMS, which surveyed a nationally representative sample of 3600 households. The headcount rate of poverty was calculated as the proportion of individuals who were consuming below the amount necessary to cover the need for basic food and non-food items. Poverty measures based on consumption are preferred to measures based on income for two reasons. First, households are less likely to report their income accurately. Second, income may vary a lot between different years, while consumption is more stable over time and therefore describes the welfare of households better. The resulting poverty headcount in 2002 for Albania was 25.4%.

However, obtaining consumption data on an annual basis is very costly and INSTAT is only collecting consumption information again in 2005. Instead, the Albanian poverty monitoring system has been designed on the assumption that poverty can be monitored using fewer data but applying statistical techniques to substitute for the lack of full information. A second round of the LSMS was conducted in 2003. About half of the households that were interviewed in 2002 were re-visited in 2003. The 2003 questionnaire included most of the questions that were asked in 2002 but did not include detailed questions on consumption. In order to assess whether the level of poverty changed between the two years, a research team from the Food and Agriculture Organisation used a methodology consisting of three steps. In the first step, all variables that were collected in both years were identified.

In the second step, the researchers assessed which variables were most closely associated with the level of household consumption in 2002. A number of variables were found to be correlated with consumption to a statistically significant degree. For example, controlling for all other characteristics, if a household possessed a washing machine then it was less likely to have low consumption. This statistical analysis estimated how strong the association of each household characteristic was with the level of consumption and the likelihood of poverty. These characteristics may change for any household from one year to another. For example, a household may have purchased a washing machine between 2002 and 2003. It may be therefore inferred that their living conditions had changed to the better.

The third step involved the use of these coefficients of correlation between consumption and household characteristics from the 2002 LSMS to predict the level of consumption using the 2003 LSMS data on household characteristics. This analysis was carried out for the national sample, as well as for the urban and rural sub-samples separately. Table 3.1 shows the results of this analysis. The first column reports the actual level of poverty calculated by the 2002 LSMS data. As mentioned above, the headcount rate of poverty in 2002 was 25.4%. The second column reports the predicted level of poverty for 2002 using the statistical model described above. This is not identical to the actual level of poverty because a prediction can only be approximate. The important comparison is between the second and the third column, which shows the predicted level of consumption in 2003 using the results of the statistical model. The estimated poverty level appears to have fallen by about 16% or three percentage points between the two years. The fall has been much smaller in rural areas.

Table 3.1Headcount poverty changes between 2002 and 2003
20022003Percentagechange
2002EstimatedEstimatedbetween 20032002and
Actual(panel sub-sample)(panel sub-sample)(%)
Total25.420.417.1-16.2
Urban19.513.210.7-18.9
Rural29.624.322.9-5.8
Source: C Azzarri et al. (2005) Monitoring poverty without consumption data: an application using the Albania panel survey, ESA Working Paper No. 05-01, Food and Agriculture Organisation
Source: C Azzarri et al. (2005) Monitoring poverty without consumption data: an application using the Albania panel survey, ESA Working Paper No. 05-01, Food and Agriculture Organisation

It is very important to stress that these are not the actual levels of poverty. A precise estimate has to wait for the results of the 2005 LSMS, which has solicited consumption information. However, the evidence in Table 3.1 can be used to argue that the poverty situation has been improving and that the benefits are more visible in urban areas. Table 3.2 translates the results of the model in terms of the predicted change in monthly per capita consumption. Compared to 2002, when the calculated average per capita consumption was Lek 7679, consumption in 2003 is estimated to have increased to Lek 8116 or by 5.7%. This is consistent with the observed rates of economic growth at the level of 6% in recent years.

Table 3.2Changes in estimated per capita consumption between 2002 and 2003
UrbanRuralTotal
Per capita consumption in 2002 (Lek per month)8,3117,0727,679
Per capita consumption in 2003 (Lek per month)8,8447,2498,116
Difference in consumption (Lek per month)533177437
Difference in consumption (%)6.42.55.7
Source: C Azzarri et al. (2005) Monitoring poverty without consumption data: an application using the Albania panel survey, ESA Working Paper No. 05-01, Food and Agriculture Organisation
Source: C Azzarri et al. (2005) Monitoring poverty without consumption data: an application using the Albania panel survey, ESA Working Paper No. 05-01, Food and Agriculture Organisation

Subjective poverty

The LSMS also asks households to assess their financial situation. Respondents are asked to imagine a ladder where the poorest stand on the lowest step and the richest on the highest step and report where they would place themselves. Such subjective measures of poverty are not perfectly correlated with objective measures of poverty. The reasons for differences vary:

People may not equate their poverty with income or expenditure alone. While objective measures relate to consumption poverty, subjective measures may also be capturing factors such as risk exposure, vulnerability, and other non-monetary dimensions of deprivation.

Quantitative measures of poverty are based on absolute poverty lines, but the way people feel about their welfare depends on their position relative to people with whom they like to compare themselves.

A recent study1 using the 2002 LSMS found that although the objective poverty status helps explain the subjective assessment to a large extent, there were substantial differences between objective and subjective welfare across households of different size. In particular, among households composed by one person, the incidence of subjective poverty is highest, while the incidence of objective poverty per capita is lowest. Households with one member are in large part (75%) old pensioners living alone: the mean age of the respondents in this group is 63 years as opposed to an average age for the total sample of 48 years. Households headed by women are also disproportionately represented in this group: 69% against an average 12% in the total sample. Also, 54% of the respondents in this group are single female pensioners; 56% are female widows; 64% suffer from chronic illnesses against an average of 28% in the total sample.

The analysis supports the argument that people’s perception of own welfare can be used to complement traditional poverty analysis. If one trusts subjective rankings, subjective poverty measurements can be a source of information for policy purposes. This makes it also interesting to observe trends. Both the 2002 and 2003 waves of the LSMS contain the same question on subjective welfare. As the following graph suggests2, perceptions have improved between 2002 and 2003. A higher proportion of respondents placed themselves among the better off in 2003.

Sources:1 G Carletto and A Zezza (2004) Being poor, feeling poorer: Combining objective and subjective measures of welfare in Albania, ESA Working Paper No. 04-12, Food and Agriculture Organisation;2 INSTAT

Protection of children’s rights

In Albania, a third of children live below the poverty line. This is a higher proportion than the average for the national population because poorer households tend to have more children. Poverty leads to poor nutrition. According to the 2000 Multiple Indicator Cluster Survey, among children under the age of 5:

  • 14% was characterised by moderate and 4.4% by significant underweight
  • 31.7% had minor and 17.3% had significant shortcomings in development
  • 11% had minor and 3.6% had significant shortcomings in nourishment

Children of mothers with secondary or higher education were less likely to be underweight. An analysis of the 3 indicators showed that the problems were more common among children aged 6- 11 months and among boys. An issue of concern remains the growth of anaemia or iron deficiency. By 2002, though, the LSMS had found that rates of malnutrition were slightly on the decrease.

Another problem that affects children’s lives is the retreat of early childhood education. In 1990, enrolment in preschool education among 3-6 year olds was 44%. This fell sharply to 28% in 1992 and has since been increasing but not recovered to the original levels (34% in 2001)1. Enrolment of children in poor families is about half that of children from well-off households. A leading reason has been the closure of public preschools, as shown below. Compared to 1990, the number of kindergartens had fallen by 60% in urban areas and by 49% in rural areas in 2004.

Poverty denies children some of their fundamental human rights. The risk to go through poverty during childhood is high in Albania. Long-term unemployment has had a negative impact on the economic and social status of families. Households headed by women are at a higher risk from poverty. About 32.8% or 293,000 children in Albania live below minimum standards2 (less than $2 per day). Furthermore, Albanian children are facing other dangers emerging during transition, such as dropouts, violence, blood-feud, trafficking, exploitation, criminality, etc.

In spite of continuous improvements in child health, infant mortality, mortality under five years of age and maternal mortality are among the highest in the region, though a decrease has been noticed in the last decade. Infant mortality has decreased from 30.8 per 1,000 live births in 1989 to 16 deaths in 20033. The rate of infant mortality and the number of underweight births indicate a deterioration of mothers’ health and nutrition. Another indicator in this respect is the increase of anaemia as result of iron deficiency.

According to a number of official documents3, a pronounced decrease of children attending preschool education has been observed in recent years. This is a result of migratory movements inside and outside the country, irregular urbanisation in some areas of the country, destruction of infrastructure and lack of security. Only 5% of the total budget of education is allocated to preschool education4.

The inheritance of a completely destroyed environment will have serious and long-term effects on children, who are particularly at risk to the effects of air, earth and water pollution. An increase in the number of deaths in children under five years of age related to conditions of the environment, such as high level of pollution and deteriorated sanitation, has been noticed, which highlights a direct link between children diseases and air pollution.

Considering some of the serious problems related to children rights and policies to be undertaken in protecting such rights, the National Strategy for Children was approved in June 2005. It is expected that the Action Plan will be approved later in June. As envisaged by the strategy, a State Committee for Children will be established to monitor and coordinate the activities of units operating in the field of children rights at the national and local levels.

The National Strategy for Children aims at establishing institutional structures and ensuring sufficient financial and human resources to accomplish obligations deriving from the Convention On the Rights of the Child. The fundamental principle in designing social policies to implement the strategy will be to ensure equal opportunities to all children, regardless of their age, sex, ethnicity, ability, status of birth etc. Priority will be given to marginalised social groups lackingparental care, exploited or ill-treated children or children with disabilities, street children and children from poor families.

To this end, a system of institutions will be created and social policy reforms will be carried out, in order to protect children from all forms of violence, exploitation and ill-treatment; to ensure living conditions for children in a family environment and, where it is not possible, to offer alternative care, giving priority to care in social families; to offer equal opportunities to children with limited abilities; to improve health care and service for mother and child; to build a modern educational system, ensuring appropriate conditions to all children to complete their obligatory education; and to establish a system of services that protect working children.

With the view of measuring the progress in implementing policies, a system of indicators and a database will be created, in harmony with the national system of monitoring the NSSED and the Millennium Development Goals. Reports will be prepared every year on the progress in implementing the strategy and the Action Plan.

Sources:1 Innocenti Social Monitor, Economic growth and child poverty in the CEE/CIS and the Baltic States, October 2004;2 Social Indicators 2004, UNICEF;3 Common Country Assessment: Albania 2004;4 Ministry of Education and Science, 2003

3.2 EDUCATION

Education is a key policy area with a well-educated and skilled workforce being central to increasing economic growth and development. The levels of education in the LSMS panel for those aged over 22 years in 2002 were relatively low with 17.7% having 4-year primary education, 45.1% with 8-year primary education, 14.7% with secondary school education, 15.1% with vocational education and 7.4% with a university level education.

Looking at enrolment rates in school for those aged 7-18 years in 2002, 75% of this age group were enrolled in school or education. Figure 3.1 shows enrolment rates by age group, poverty status of the household and urban or rural area. Those aged 16 - 18 years were least likely to be enrolled (35%) compared to 85% of 11-15 year olds and 95% of those aged 7-11 years. Children from poor households were less likely to be in school or education than children from non-poor households. Children in urban areas were more likely than those in rural areas to be enrolled in school or education (85% in urban areas compared to 69.5% in rural areas).

Figure 3.1Enrolment status of children aged 7 - 18 years, 2002

Even though children in rural areas were less likely to be enrolled in school, the effect of the poverty status of the household is more marked in urban areas than in rural areas. Of children in poor households in urban areas, over one quarter (26%) were not enrolled in school compared to 17% of non-poor children in urban areas, a percentage point difference of 9%. In rural areas, non-poor children were only 4% more likely to be enrolled in school than poor children.

Those who were not enrolled in school were asked for the reason. The main reason given was that they had completed their studies or had no interest in continuing in education (54%), followed by having other work to do, including agricultural work (16%) and the school being too far way (15%) even though this last reason applied in rural areas only. Of those who were not enrolled in 2002, 50% were working as farm workers, 5.5% were employees, 2% were selfemployed, 29% were unemployed and 13.5% had some other status e.g. ill, housewife.

Dropout rates as children become older are of concern. The key ages where the risk of dropping out is greatest are from 13 to 16 years. In 2002, 6% of 13-year olds were not in school, increasing to 23% at 14, 42% at 15 and 60% at 16 years of age.

Changes between 2002 and 2003 for 15 - 18 year olds

Using the panel component of the LSMS for 2002 and 2003 the dropout rates for 15 - 18 year olds who were in education in 2002 can be seen. Of those who were in education in 2002, 86.4% were enrolled at both years while 13.6% were not enrolled in education by 2003. The dropout rate varies by age with those aged 15 in 2002 being most likely to drop out by 2003 (25%) compared to 10% of those aged 16 in 2002 and 6% of those aged 17 in 2002. So the dropout rate is greatest between the ages of 15 to 16 years, something which is in line with the legal age for leaving formal education and the typical completion age of 8 years in primary education.

3.3 HEALTH

Respondents on the LSMS were asked to rate their own health. Overall, the majority of respondents (69%) rated their health as being very good or good, 19% said it was average and 12% rated their health as being poor or very poor. There are some differences between men and women, with men being less likely than women to rate their health as average or poor/very poor. Of those rating their health as poor or very poor, 35% were men and 65% were women. Self-rated health status also varies by age group with the older age groups being more likely to rate their health as poor or very poor. Table 3.3 shows the self-rated health status reported by respondents aged 15 years and over in 2002. Those with a chronic illness or disability were significantly more likely to rate their health as poor or very poor (85% compared to just 15% of those with no chronic illness).

Table 3.3Self-rated health status of individuals aged 15 and over, 2002 (%)
Health status
Very good/

Good
AveragePoor

/ Very poor
Total number

of observations
Suffers from chronic illnessYes4.245.385.41095
No95.854.714.63911
Housing conditionsGood89.487.084.54417
Poor10.613.015.5582
Poverty statusNon-poor77.778.974.73865
Poor22.321.125.31134
All68.819.411.8100
Number of observations34409725945006

At each year of the panel survey respondents have been asked to rate their housing conditions according to whether they are very good, appropriate for living, inappropriate for living and under construction (mostly incomplete). Those living in poor housing conditions were more likely to report having poor health. Of those saying they had poor health 15.5% were living in poor housing conditions compared to 11% reporting good health. There was also an association between poverty and poor health with 25% of those with poor health living in a poor household compared to 22% of those who had good health.

Health status in urban and rural areas

Households were asked about their main source of water according to whether they had running water inside the dwelling, running water outside the dwelling, used a water truck, public tap, spring or well, river lake pond or some other source. They were also asked if they had a toilet inside the house, 2 or more toilets inside the house, a toilet outside the house, or some other type of toilet.

There were some clear differences between health status and housing conditions in urban and rural areas. On the whole, people living in urban areas reported better health and also having better housing conditions than those living in rural areas. Looking first at urban areas, 69.5% of those in good housing conditions rated their health status as very good/good compared to 62% of those in poor housing (Figure 3.2). A similar pattern can be seen with indoor running water and indoor sanitation, with those having these facilities reporting better health.

Figure 3.2Self-rated health status by housing conditions in urban areas, 2002

Looking at rural areas, there are some marked differences compared to urban areas. In total, 16.5% of rural households were living in poor housing conditions compared to 6% in urban areas. In rural areas, 23% of respondents were living in a house with no indoor water compared to 12% of respondents in urban areas. Just 38% of people in rural areas were living in a house with an indoor toilet compared to 86% of those in urban areas.

In rural areas, 59% of respondents living in good housing conditions had good health compared to 53% in poor housing conditions (Figure 3.3). In the case of rural areas, those who had poor health were also more likely to have no indoor water or indoor toilet but the differences are greater than in urban areas. Of those with no indoor water, 18.5% rated heir health as poor/very poor compared to 12% of those with indoor running water. Similarly, those with an indoor toilet rated their health better than those with some other type of toilet.

Figure 3.3Self-rated health status by housing conditions in rural areas, 2002

Usage of health services

Respondents were asked how many times they had used a public ambulatory service, visited a private doctor or seen a nurse, paramedic or midwife in the past 4 weeks. They were also asked how many times in the past twelve months they had been admitted to hospital and been to a dentist. In total, 38.8% of all respondents had used at least one of these services when surveyed in 2002 and 36.5% had used at least one in 2003. In 2002, 23.5% of respondents had gone to the dentist in the past twelve months but in relation to services used in the past 4 weeks the public ambulatory service was the most frequently used by all respondents, with 15.7% of people using this service.

In 2003, the results are similar, with 22.3% of respondents visiting the dentist in the past twelve months and 11.3% using the public ambulatory service in the past 4 weeks. Table 3.4 gives the mean number of visits to each of these types of services. In 2002 the service with the highest mean number of visits in the past 4 weeks was paramedic or midwife, 5.78 visits for those using this service and 0.16 for the whole population. However, relatively few respondents used this service which suggests high usage by a limited group of people. The public ambulatory had the next highest mean usage at 1.54 visits in the past 4 weeks for those who used it. This was also the service most used by the whole population with 0.24 visits on average.

Table 3.4Mean usage of health services, 2002 and 2003
20022003
Type of service usedMean number of visits by the populationMean number of visits by individuals who used the serviceMean number of visits by the populationMean number of visits by individuals who used the service
In past 4 weeks
Public ambulatory0.241.540.161.36
Private doctor0.031.450.031.38
Nurse / Paramedic / Midwife0.165.780.114.62
Outpatient visit to hospital0.01NA0.071.52
In past 12 months
Admission to hospital0.061.280.061.25
Dentist0.662.790.562.48

Those living in poor households had a lower mean usage of health services even though the differences were not substantial compared to non-poor households, a pattern which was similar in both 2002 and 2003. In general, the mean usage of health services in 2003 was slightly lower than in 2002 even though the differences are not great.

Smoking

In 2003 respondents were asked whether they smoked cigarettes and smoking is associated with reporting having worse health. The data are reported for men only as the percentage of women reporting smoking in the survey was very low and may not be reliable. It may be the case that women are more reluctant to admit they smoke, as it is not seen as being socially desirable or acceptable for them to smoke. Almost one third of men smoke (31%). Smokers were less likely to report having good health than non-smokers with 69.6% of smokers having good health compared to 76.1% of non-smokers.

Table 3.5Relationship between smoking and self-rated health status of men, 2003
Health statusNon-smokerSmokerTotal number of observations
Very good / Good (%)76.169.61703
Average (%)15.319.8385
Poor / Very poor (%)8.610.6217
Total number of observations15797262305

3.4 LABOUR MARKET

Labour force participation

The labour force in Albania was calculated to be 1.3 million in the 2001 Census or lower by 15% since 1989. The workforce lost more than a quarter of women and 5% of men. In 1989, there were only slight differences in labour force participation between men and women. By 2001, gender equality had disappeared, as only half of the adult women were active in the labour market and the proportion of inactive women had doubled. Labour force participation among men remained roughly the same but its composition changed as the share of the employed fell (61% relative to 71% in 1989) and the share of unemployed rose (14% relative to 6%). The size of the male labour force was reduced in absolute numbers as a result of emigration.

Table 3.6Labour force participation according to the 1989 and 2001 Census
19892001
MaleFemaleTotalMaleFemaleTotal
In thousand
Employed77966414436503921042
Unemployed6294157150155306
Total active84275816008005481347
Inactive249283532269554823
Total (aged 15 years or more)109010412132106911012170
As percentage
Employed71.563.867.760.835.648.0
Unemployed5.79.07.314.014.114.1
Total active77.272.875.074.849.762.1
Inactive22.827.225.025.250.337.9
Total (aged 15 years or more)100100100100100100
Source: People and work in Albania, Population and Housing Census Research Publication, INSTAT, 2004
Source: People and work in Albania, Population and Housing Census Research Publication, INSTAT, 2004

It is important to keep in mind that the notions of employment and unemployment have changed since 1989 and are not easily comparable. However, some changes are unambiguous, such as the drop in the female participation rate. Despite the high emigration rate of men, it appears that women have not substituted them in the domestic economy, not even in rural areas or agricultural activities. Female employment decreased mostly in urban areas for several reasons: the closure of factories, the increasing significance of male-oriented activities, such as trade and construction, and internal migration to urban areas, as women face more problems entering the urban labour market in the current economic environment.

Employment

Measuring employment is fraught with difficulties in the context of a Census. Many people do not consider activity in the informal sector as equivalent to being employed. Respondents have sometimes an incentive not to report their real employment status. For example, according to the Census, the employment rate of 46-year-old women jumps by 20 percentage points relative to the rate of 45-year-old women. One of the likely explanations relates to social insurance rules for former employees of agricultural enterprises and members of their families. To enjoy a pension starting at age 55, women must have been employed for at least 10 years. Many women believed that claiming to be employed would be useful to earn a pension.

The direct and probing interviews used in the LSMS were designed to capture the employment status of the population better than the Census. Table 3.7 shows that men are more likely than women to be in employment (55% of the employed were men). At the same time, women are more likely than men to be inactive (69%) or unemployed (53%). Of those aged 15-34 years, 54% were in employment, while of those aged 35-54 years 52% were in employment. The unemployed were most likely to be aged 15-34 years (36.6%) and those who were inactive were more likely to be in the younger age group of 15-34 year olds.

Table 3.7Employment status of working age individuals, 2002
EmployedUnemployedInactiveTotal number

of observations
Gender
Male54.946.830.91997
Female45.153.269.12176
Age
15-3454.013.232.81901
35-5452.436.627.31874
55-over9.12.819.6398
Highest qualification
Up to 8 years57.658.669.52179
Secondary / Vocational31.536.826.71170
University10.94.63.8309
Total61.716.022.3
Note: Working age is defined as 15-64 years for men and 15-59 years for women.
Note: Working age is defined as 15-64 years for men and 15-59 years for women.

Of those who were in employment, the majority (58%) had up to 8 years education, 31% had a secondary or vocational education and 11% had a university qualification. This reflects the relatively low levels of educational qualifications in the population as a whole and is in line with the fact that most people work in agriculture (53%) and construction (10%). According to the 2001 Census, about a third of all jobs held were temporary, seasonal or occasional.

Unemployment

One of the key differences between the Census and the LSMS assessments of the labour market is the fact that it resulted in a lower unemployment rate for women, as many women were not systematically searching for a job. The LSMS is better placed to distinguish the boundaries between discouraged unemployment and inactivity.

The trend in unemployment has been downward between 2002 and 2003 according to the LSMS. INSTAT uses 2 definitions. The standard definition of unemployment requires that the individual has actively engaged in some type of job search. The relaxed definition of unemployment also includes seasonal workers or discouraged job seekers, who may be considered as unemployed even though their attachment to the labour force is marginal. Under both definitions, the unemployment rate appears to have fallen by 2 percentage points between the 2 years, which is a substantial achievement. This improvement in employment indicators ties in with the current rates of economic growth.

Figure 3.4Unemployment rate according to the standard and relaxed definitions

Source: INSTAT calculations with the 2002 and 2003 LSMS data

The LSMS design, which interviews the same individuals each year, allows a comparison of the employment status across years for the same respondents. Table 3.8 shows the employment status for individuals of working age interviewed in 2002 and 2003 and shows that there is a quite a lot of movement between employment states. The percent highlighted in bold on the diagonal shows those who were in the same employment status at each year of the survey. The percentage off the diagonal shows those who had changed their employment status by 2003.

Table 3.8Employment status of working age individuals, 2002 and 2003 (%)
2002
EmployedUnemployedInactive
2003
Employed87.539.620.6
Unemployed4.839.311.2
Inactive7.721.168.2
100100100

The data suggest that for those who are in employment their status is relatively secure: 87.5% of these people were in employment at both years of the survey. Those who were unemployed in 2002 were the least stable category with just 39.5% giving the same response at both years. On the contrary, 39.6% had moved into employment by 2003 while 21.1% had become inactive, providing some positive evidence that over one third of the unemployed had found employment within a year.

Those who were not working were asked about their job search behaviour and how they had gone about looking for work. When people were asked in 2002 about their efforts to find a job, 64% of them said that they had looked for work through friends or a relative, 26% through a labour office and 4% through contacting the employer. The length of time people spend looking for work varies but for many people it is a long process. On average, 32% of respondents who were not in work in 2002 had been looking for work from between 12 to 23 months and at 2003, 42% of respondents had spent up to 2 years looking for work. With relatively long periods of unemployment, there is a danger that some people may stop trying to find work.

Investing the earnings from international migration

The decision to migrate is considered the result of a joint decision by members of a household and assumes that the household members share the costs and benefits of this decision. This explains the motivation of migrants to remit earnings to household members.

Earnings from migration do not only return to Albania in the form of remittances by household members who reside permanently abroad but is also brought back by temporary migrants who spend limited parts of the year outside the country. Short-term migration is a less well known phenomenon. Evidence on it is provided by the 2002 LSMS. The survey included information on the migration history of individuals who had been abroad for at least 3 months at any time between 1997 and 2001 but were in Albania at the time of the survey. A fifth of households have members who have been temporary migrants, whereas a third of households have members who are permanent migrants.

A study based on these data1 shows that households with temporary migrants are larger than the average Albanian household. Households whose members migrate temporarily to Greece tend to live in rural areas of the central and mountain zones, are poorer and less educated. Households whose members migrate temporarily to Italy tend to originate from the rural areas of the coastal zones, are better off and more educated than the average household.

Characteristics of households

with temporary migrant members
No

migrants
Migrants

to Greece
Migrants

to Italy
Household size (members)4.24.74.5
Average education (years)8.68.49.4
Monthly consumption per capita (Lek)784170199205
Living in rural areas of the coastal zone (%)80137
Living in rural areas of the central zone (%)77203
Living in rural areas of the mountain zone (%)76195
Proportion of households (%)82135

An important policy question is how to steer the income that migrants, whether temporary or permanent, bring back to the country. If income from migration is invested in productive activities, then it could help create jobs in these rural areas and perhaps reduce the flow of migration. A study that used the results of a specific survey in the Korca region2 examined the effects of migration and remittance income on the productivity of agricultural households. The results showed that households without migrant members were less efficient in farming. A potential explanation is that the recipients of assistance lower their effort in the knowledge that remittance income guarantees a certain level of welfare.

Although remittances may well be used for investment outside agriculture, complementary evidence from the LSMS confirms that they are most commonly used to pay for basic necessities and durable goods. The government has been formulating a migration strategy with support from the International Organisation of Migration, which will also include actions to facilitate the flow of remittances into business activity. Building confidence and increasing the efficiency of the banking sector is considered a necessary step3.

Sources:1 G Carletto et al. (2004) Internal mobility and international migration in Albania, ESA Working Paper No. 04-13, Food and Agriculture Organisation;2 E Germenji and J Swinnen (2004) Impact of remittances on household-based farms in rural Albania, Paper presented at the conference on ‘New perspectives on Albanian migration and development’, Korca, September 2004;3 E Uruçi and I Gedeshi (2003) Remittances management in Albania, Centro Studi di Politica Internazionale Working Paper 5/2003

3.5 WATER AND ELECTRICITY

Access to basic facilities such as water and electricity are indicators of households’ quality of life and as noted in the health section earlier, those who did not have access to running water and sanitation within their house tended to report worse health.

Water supply and quality

Of all households in 2002, 50% had running water inside the dwelling, 17% had running water outside the dwelling, 22% used water from a spring, well or river while 11% used a public tap or water truck. Just under one third of households (32%) had a continuous supply of water and poor households were less likely than non-poor households to have running water or a continuous supply.

Figure 3.4 presents indicators of access and quality of water supply in urban and rural areas in 2002 and 2003. Urban households were more likely than rural households to have running water inside their dwelling. In contrast, rural households, where water was more likely to be from a spring or well, reported higher water quality levels in the sense that it was more often suitable for drinking. Similarly, those in rural areas were less likely to say that they boiled their drinking water than those in urban areas.

Figure 3.4Water source and quality, 2002 and 2003

Between 2002 and 2003, there were improvements in water indicators, although the inequities between urban and rural areas remained broadly the same. Overall, there was an increase in the proportion of households with running water inside their dwelling to 60%. Conditions improved for rural areas where the proportion of households with running water inside the house increased from 24% in 2002 to 33% in 2003. In 2003, rural areas were as likely as urban areas to have a continuous water supply and the continuous supply had improved in all areas. The proportion of households reporting that their water was not good for drinking increased slightly in 2003 but on the other hand the percentage boiling their drinking water fell in both urban and rural areas. However, this may in part be because more households used bottled water for drinking (an increase from 40.3% of households in 2002 to 44.1% in 2003).

For households with running water either inside or outside the house, the average supply fell from 5.5 to 4.9 hours per day between 2002 and 2003. However, the mean for all households increased slightly from 1.9 to 2.0 hours per day during the same period. Urban areas were supplied with water 3 hours per day on average, while in rural areas the average supply was just 1 hour.

Electricity

Improving the electricity supply is considered a priority, in particular to improve the regularity of the supply and reduce the number of electricity cuts experienced by households and businesses. In 2002, 13.5% of households in the LSMS said they never had any electricity cuts, 5.6% had cuts several times a month, 8.6% had cuts several times a week and 77.2% had cuts every day. For those experiencing cuts, the mean number of hours per day without electricity in the month before the survey was 8.6 hours. Urban households had on average no electricity for 7.2 hours per day and rural households for 9.5 hours per day.

In 2003, the mean number of hours households had no electricity had fallen to an average of 4.8 hours per day. In urban areas the mean was 3.5 hours and in rural areas 6 hours per day. In 2003 households were also asked whether they thought the electricity service had improved over the past year. The majority (52%), thought that the service had improved over the past year while a further 28.5% thought it had stayed the same. Only 17.5% thought that the electricity service was worse than the year before. Figure 3.5 shows the distribution by urban and rural areas. Households in urban areas were more likely than those in rural areas to say that the electricity service had improved in the past year while those in rural areas were more likely to say that it had remained the same or worsened.

Figure 3.5Perception of the electricity service quality between 2003 and 2002

Other Resources Citing This Publication