PORTRAITS OF POVERTY AND INEQUALITY IN ALBANIA Portraits of Poverty and Inequality in Albania Based on Census 2011 and Living Standards Measurement Survey 2012 2016 Table of Contents Abbreviations vi Preface vii Team and Acknowledgements ix 1 Introduction 1 2. Constructing the poverty maps for Albania 5 2.1 Methodology 5 2.2 Data 6 The 2011 census 6 The 2012 Living Standard Measurement Study 7 2.3 Model 7 The consumption models 7 Selecting the model variables 8 The stability and accuracy of the consumption models 9 Standard errors and the level of aggregation 10 3 Results 12 3.1 General analysis of the results of the mapping exercise 13 Prefectures 13 Municipalities and communes 13 4 Using poverty maps in policy making 16 4.1 The poverty maps for policy in Albania 16 Poverty maps of the new local administrative units 18 Overlaying poverty maps with other data of policy interest 21 Spotlight 1. Poverty maps and policy making around the world 23 5. Conclusions 26 References 30 Results of the 2012 Poverty Maps 33 Results of the 2012 Poverty Maps - New Municipalities 42 Annex C: Explanation of the Variables and Relevant Data Tables 44 Additional Annexes Main Annex - A closer look: results by prefecture 48 Berat Prefecture 48 Dibër Prefecture 51 Durrës Prefecture 54 Elbasan Prefecture 57 Fier Prefecture 60 Gjirokastër Prefecture 63 Korçë Prefecture 66 Kukës Prefecture 69 Lezhë Prefecture 72 Shkodër Prefecture 75 Tirana Prefecture 78 Minimunicipalities of Tirana 81 Vlorë Prefecture 84 Annex A: Updates to the First Albania Poverty Maps, 2001–2010 87 Annex B: The ELL Methodology: Innovations and Criticisms 88 Annex D: Differences and Similarities: 2002 and 2012 Maps 90 Abbreviations ELL Elbers-Lanjouw-Lanjouw (poverty mapping method) INSTAT Albanian Institute of Statistics LSMS Living Standards Measurement Survey PSU Primary Selection Unit vi Portraits of Poverty and Inequality in Albania Preface The Albanian Institute of Statistics (INSTAT) and the World Bank have developed the following report on poverty mapping in Albania. Small area poverty maps portray subnational geographical areas, such as municipalities, counties, and districts, and highlight the small areas most likely to exhibit the highest risk of poverty in a country. The poverty maps displayed in this report are designed to help policymakers, development partners and other stakeholders identify the parts of Albania with pockets of poverty, to inform targeted efforts to achieve the goals of reducing poverty and inequality. The report describes the technical aspects of the process of creating — based on Census 2011 and Living Standards Measurement Survey 2012 data — accurate, detailed geographical presentations of poverty in the country and summarizes the nature of the insights about poverty and shared prosperity that are revealed by the maps. It displays sample maps to aid the reader in understanding the value of the approach by highlighting that the more finely grained poverty maps produced through the project expand the evidence supplied through more traditional and less-detailed geographical representations of poverty. Portraits of Poverty and Inequality in Albania vii Team and Acknowledgements This report has been produced jointly by the Albanian Institute of Statistics and the World Bank, under the guidance of María E. Dávalos (Senior Economist, World Bank) and Ledia Thomo (Director of Household Surveys, Albanian Institute of Statistics), with main contributions from Gianni Betti, Laura Neri, Thomas Pave Sohnesen, Robert Zimmerman and the Albanian Institute of Statistics technical team, particularly Teranda Jahja, Ruzhdie Bici, Blerta Muja, Alma Kondi and Mirela Deva. The team is greatful for comments received from Alexandru Cojocaru and Willian Seitz from the World Bank. Supervision of this work has been provided by Mirela Muca (Director General, Albanian Institute of Statistics), Gjergji Filipi (former Director General, Albanian Institute of Statistics), Carolina Sanchez-Paramo (Practice Manager, Poverty and Equity Global Practive, World Bank) and Tahseen Sayed (Country Manager for Albania, World Bank). Portraits of Poverty and Inequality in Albania ix CHAPTER 1 Introduction As long as poverty and social exclusion exist, social progress may only advance haltingly because some population groups are unable to participate fully in the economic and social life of a country. Accurately identifying the poor and where they are located may thus help in delivering antipoverty relief and basic services and guiding programs aimed at reducing poverty and enhancing the equitable sharing of prosperity. Moreover, funding and targeting basic services and programs to reduce poverty and social exclusion within a country depend on the availability of detailed knowledge of the disparities in well-being across households. Traditional methods of identifying the distribution of poverty across regions, provinces, and smaller administrative divisions rely on household survey data. However, household surveys are limited in their ability to localize the poor at low levels of geographical disaggregation and thus circumscribe the information upon which policy makers may rely. Poverty is most often measured using data on consumption from a sample survey, such as the Living Standards Measurement Survey (LSMS), in which household per capita expenditures (or per adult equivalent expenditures) are compared with a fixed poverty line. Given the complexity and cost of undertaking this or a similar exercise, it is only feasible to collect this information from a sample of households in a survey. However, data describing the distribution of income across households at a high level of geographical aggregation often mask critical information from policy makers. For instance, there may be large numbers of the poor concentrated in small discrete areas in regions that show low overall poverty rates. However, poverty measures based on surveys have sampling errors, and these rise rapidly if we calculate such measures for smaller areas. This precludes analysis of poverty in such areas in this way. This study is based on 2011 Census and 2012 LSM and this shortcoming exists for the LSMS in Albania, as well as household surveys across most countries, despite efforts to improve the 2012 round of the study to provide poverty data that are representative of the 12 prefectures in the country. Poverty mapping has emerged as a possible way around this problem. Poverty maps allow a focus in greater detail on the spatial distribution of poverty at the local level. As a result, reasonably precise poverty estimates can be obtained through this technique. The government of Albania has embraced the approach. Indeed, Albania has been a pioneer in applying the poverty mapping methodology. In 2001, the World Bank, in collaboration with the U.K. Department for International Development, began assisting the government in a project to establish a permanent poverty monitoring mechanism. The objective was to create a reliable, sustainable system for the timely production of statistical information to guide decision making and policy making. It was expected the method would Portraits of Poverty and Inequality in Albania 1 Map 1: Poverty Maps and Poverty Rates (Headcount Ratio), Albania, 2002 a. Communes, municipalities Source: INSTAT 2004. be particularly useful in the design, implementation, and evaluation of economic, social, and environmental development and financing programs so that these could efficiently target the areas of the country with the greatest need. As of late 2000, Albania did not yet have a reliable, nationally representative survey to assess the living conditions of the population. In 2001, however, the Albanian Institute of Statistics (INSTAT), with donor support, carried out the country’s first census since the end of the communist era. The following year, INSTAT implemented a nationally representative multidimensional household survey. That same year, the World Bank began to prepare a poverty assessment, and the government completed the preparation of a Poverty Reduction Strategy Paper. The convergence of these events and the related data collection efforts supplied 2 Portraits of Poverty and Inequality in Albania the ideal opportunity to construct the first small area estimates of poverty and inequality in Albania, as well as poverty maps (see annex A, available online). The maps clearly show how an analysis of poverty might improve because of the detail at a more finely disaggregated level (map 1). Using the 2005 and 2008 LSMS, new rounds of the maps were calculated still based on the 2001 census. The aim of the creation of the poverty maps was to estimate poverty and inequality for each of the prefectures, districts, and municipalities or communes, and for each minimunicipality of Tirana. This was the goal because (1) the poor in Albania are concentrated in rural and mountainous areas, but also in areas that are more well off; (2) the government was seeking to decentralize the delivery of services to local governments, which might develop independent strategies and interventions; and (3) the targeting of government antipoverty programs needed to be improved (World Bank 2015). All these aims show useful possible applications of the maps. Following the mapping exercise, individuals involved in the preparation of the maps and potential users of the maps were interviewed. The interviews revealed three ways in which the poverty maps were applied in decision making and policy making: (1) as a benchmark against existing resource allocation criteria, for example, whether the allocation of social-assistance block grants according to previously established criteria correlate with an appropriate allocation based on current poverty rates; (2) as a tool in targeting public spending; and (3) for the provision of data to monitor the progress toward achieving the Millennium Development Goals (World Bank 2015). Several nongovernmental organizations also relied on the poverty maps in supplying advisory services to local governments and donor agencies and in designing joint intervention strategies. The poverty maps were likewise used to improve the prioritization of investments in secondary roads. This report presents results for the new poverty maps that have been created in Albania using data of the 2012 LSMS and the 2011 Population and Housing census and incorporating improvements to the poverty mapping technique. This opens up a new opportunity to, with the most recent data available on welfare in Albania, better understand the distribution of poverty and inequality across the country, identify pockets of poverty and inform policymaking. The report consists of five sections, including this introduction. The next section describes the process for constructing a poverty map. The following section outlines the main results of the new poverty mapping in Albania. With the aim of stressing the impact that poverty maps can have on policy making, the subsequent section highlights the applications of poverty mapping in policy making in Albania and elsewhere. The final section has the conclusions. There are also included five annexes, some available only online as indicated. Annexes provide more detail on the full results, selected technical aspects of the creation of poverty maps in Albania, and more precise information on the variables included in the models of consumption used in the construction of the maps. Portraits of Poverty and Inequality in Albania 3 CHAPTER 2 4 Portraits of Poverty and Inequality in Albania Constructing the poverty maps for Albania A poverty map is a geographical profile in map form that shows the distribution of poverty within a country. It is used by policy makers to plan targeted public investments in education, health care, sanitation, water, transport, and so on to reduce poverty and social exclusion and thereby promote shared prosperity. Relevant institutions engaged in social issues and the development of social policies often use such maps so that investments designed to lower poverty rates can reach a maximum of the poor in a country or in smaller areas within a country. A poverty map is therefore most useful if it can be constructed at a detailed level of geographical disaggregation, such as states or provinces, counties or districts, and municipalities/communes, while also highlighting the small areas most likely to exhibit the highest risk of poverty and social exclusion and supply a comprehensive picture of the spatial distribution of poverty and social exclusion. However, achieving this outcome requires large datasets, which are typically provided only in nationwide household surveys and population and housing censuses. Yet, a household survey cannot both cover a large household sample and provide detailed information on household welfare, such as precise data on household expenditure or income. Typically, these surveys rely on sampling but are able to collect a wide range of rich data, from household expenditure to nonmonetary indicators, for example, assets, access to basic services, the physical quality of the home, and so on. Meanwhile, censuses do not generally collect detailed information on income or expenditures because the cost of gathering such data across an entire country is exorbitant. In both household surveys and censuses, there is a trade-off between survey size and data detail because achieving either goal is costly in terms of money and manpower. 2.1 Methodology Elbers, Lanjouw, and Lanjouw (2002, 2003) have developed a method to create poverty maps by merging a household survey and a census to take advantage of the strengths of the two sources of data while minimizing the weaknesses. The strength of the household survey is the extensive detail on household characteristics that can be obtained because the samples are relatively limited. In the case of poverty, the basis might be, for example, information on household consumption. The strength of the census is the universal or near universal coverage of households. The method has become widely popular among development practitioners around the world. The method proposed by Elbers, Lanjouw, and Lanjouw (ELL) has three stages. In stage 0, a set of variables common to both the survey and the census are identified and evaluated in terms of the extent to which their distribution across the two data Portraits of Poverty and Inequality in Albania 5 sources is similar. In stage 1, a model of log per capita consumption expenditure is estimated in the survey data based on the identified variables. The model includes household-level variables that have a similar distribution in both the survey and the census. The model also includes the location-specific averages of variables found in the census and, potentially, other external variables available across small local levels across an entire country, such as variables pinpointed spatially through a geographic information system. In stage 2, poverty estimates and the associated standard errors are computed. A simulated value of consumption for each census household is calculated, along with the predicted log expenditure and random draws from the estimated distributions of the disturbance terms. The ELL method offers a way to calculate poverty estimates and the standard errors properly, while accounting for potential sources of bias.1 2.2 Data Beginning with the exercise in 2002–04, poverty mapping in Albania has been based on the ELL method. The poverty maps produced then have now been updated using two primary data sources, the most recent (2012) LSMS and the most recent (2011) population and housing census. The updating has relied on several new methodological developments, including suggestions of Elbers and van der Weide (2014). (See annex B, available online for innovations in and criticisms of the ELL methodology.) The comparability of the two rounds of poverty maps is not straightforward (considerations in Annex D) and has not been conducted in this report. The 2011 census The last population and housing census, that of October 1, 2011, recorded 2.8 million residents in Albania. The census included a set of variables that can be compared with the variables of the LSMS 2012. Compared with the previous (2001) census, the 2011 census found notable changes. Thus, between 2001 and 2011, the population fell by 269,000 (8.8 percent); the average age of the population increased from 30.6 years to 35.3 years; the number of under-15-year-olds sharply declined, from 898,000 to only 579,000; and the number of people 65 years of age or older rose from 231,000 to 318,000. Between the 2001 and the 2011 census, 228,952 individuals changed their prefectures of usual residence. These migrants accounted for 8 percent of the resident population in 2011. During the decade, there were 280,863 migrants between towns or villages. This internal migration led to large-scale urbanization 1. There are two sources of error involved in the estimation process: (a) errors in the estimated regression coefficients and (b) errors in the disturbance terms. Both types affect poverty estimates and the accuracy of the estimates. 6 Portraits of Poverty and Inequality in Albania in some areas and substantial depopulation in others. Because of the economic and logistic difficulties facing migrants seeking to settle in cities such as Durrës and Tirana, the surrounding areas received a wide majority of the internal migrants. Nearly half relocated to Tirana Prefecture. The 2012 Living Standard Measurement Study INSTAT conducted the LSMS in 2002, 2005, 2008, and 2012 to study various socioeconomic characteristics of the population, including consumption and poverty. The objectives of these surveys were similar. However, while the geographical domains in the 2002, 2005, and 2008 LSMS were defined as the urban and rural strata in four regions (central, coastal, mountain, and Tirana), the domains in the 2012 LSMS, which was carried out in September and October 2012, were expanded to produce representative results by urban and rural strata in the 12 prefectures of the country. This required a considerable increase in the sample size. The data of the 2011 census provided the sample frame for the 2012 LSMS. This was considered a stratified, two-stage cluster sampling design. During the first stage, the sampling units were 834 primary selection units (PSUs) selected among the census enumeration areas, while, in the second stage, the sampling units were the households in each selected PSU. The selection of the households within each chosen PSU followed the procedure applied in the 2002, 2005, and 2008 LSMS. In this procedure, an initial sample of 12 households is selected according to the systematic sample. Using systematic sampling, 8 of these are then selected as the base sample, while the remaining 4 are substitutes that were available for use in case households in the base sample could not be contacted or were nonresponsive. The instruments for the collection of information in the 2012 LSMS were (1) a household questionnaire, (2) a diary recording household food consumption, and (3) a price questionnaire. The final 2012 sample consisted of 6,671 completed questionnaires. 2.3 Model The poverty mapping methodology requires, in stage 1, the estimation of a model of consumption based on the household survey. This model serves as the basis for estimating the consumption of each household in the census during stage 2 of the process. This subsection details how the model was constructed for the new round of poverty maps in Albania. The consumption models The Albanian poverty map is based on four regional consumption models (central, coastal, mountain, and Tirana). Why four regional models instead of one national model? The short answer is that these four domains were defined for sampling in Portraits of Poverty and Inequality in Albania 7 the LSMS because they were internally homogenous, thereby optimizing the sample size. This homogeneity makes these domains ideal for modeling consumption. There are advantages and disadvantages to splitting the LSMS sample into four models. The division helps in more effectively capturing local circumstances, but also decreases the number of observations in the four domains or regions, thereby limiting the number of variables that can be included accurately in each model. The implicit assumption is that the parameter estimates on the regression are the same across households in each domain. Meanwhile, a national model assumes that the relationship between household expenditure and household characteristics is uniform throughout the country. This may be a much less tenable assumption. Using separate models that are more internally homogenous allows the relationship between expenditure and the explanatory variables to vary by region, and it reduces the standard error of poverty prediction because of an error in modeling. Nonetheless, if the domains chosen for the models are too small, they may become prone to overfitting, and the predictions could become overly influenced by idiosyncrasies in the LSMS sample. Hence, there needs to be a balance between heterogeneity across the country and a smaller, but still adequate sample size. In Albania, the four domains seem to achieve this because the models replicate poverty as measured by the LSMS, without showing relatively large variations in the cluster effect, and they are based on relatively parsimonious models with high R squared values (see annex C, tables C.2 and C.3). Selecting the model variables As laid out in the methodology subsection above, a number of steps are required to create a poverty map. This subsection describes the work and the checks on the data in terms of the alignment of the explanatory variables in the 2011 census and the 2012 LSMS. The definition of poverty is the same in the poverty maps and in the LSMS survey, and the national poverty line of 4,891 ALL (in constant 2002 new leks) has been applied to all results (INSTAT 2013). In the poverty maps, only variables that show similar distributions in the LSMS and the census are eligible as explanatory variables in the regression models. The estimation of the Albanian poverty maps represents a nearly perfect setting for this procedure because the census includes many variables that are highly correlated with consumption, including the following: • Demographic characteristics: sex, age, marital status, household size, number of children, number of adults, and number of the elderly in the household • Education: educational attainment of the household head and the highest 8 Portraits of Poverty and Inequality in Albania educational attainment achieved by any household member • Occupation: employment status, occupation, and sector of employment. • Housing characteristics: type of housing unit, age of the building, the presence of an elevator, the presence of a toilet, the source of water, the number of rooms, the size of the dwelling, dwelling ownership, and occupancy status. • Durable goods and productive assets: possession of a furnace, refrigerator, freezer, television, television decoder, washing machine, dishwasher, microwave, fixed telephone, mobile telephone, computer, Internet access, solar panels, air conditioner, automobile, and agricultural land and livestock. Furthermore, the 2012 LSMS and the 2011 census were conducted at approximately the same time (within a single year), thereby limiting the possible shifts in the variables because of changes over time. The full list of questions from each data source—the LSMS and the census—demonstrates the degree of comparability. Some variations in the questionnaires used for the LSMS and the census may produce variations in the distribution of the variables. A difference in questionnaire design in terms of length and complexity may also drive variations in reported values even if the questions are similar (Kilic and Sohnesen 2015). (The 2012 LSMS questionnaire was longer and more complex than the 2011 census questionnaire.) Other characteristics of a survey or census, such as the training of enumerators, the characteristics of field implementation, and so on, may also have an impact on the data collected. Variables were considered for inclusion in the final models for the mapping exercise only if their distribution, domain by domain, was found to be similar in the survey and the census. (Annex C, table C.1 shows the survey and census mean for the variables included in the models.) The stability and accuracy of the consumption models A good predictive model needs to balance several objectives. It should reveal a high correlation between consumption and household characteristics. This can be gauged through the adjusted R squared of the consumption regression. However, maximizing the R squared can easily generate weaknesses. One potential weakness is overreliance on the specific survey sample. To avoid this and other weaknesses, the mapping models have been designed not to be too specific to the selected LSMS sample. This has been accomplished by testing different models and comparing final predictions and by excluding variables with skewed distributions, that is, variations relying on relatively few observations. Furthermore, the variables found to have the greatest importance in random forest predictions are considered the Portraits of Poverty and Inequality in Albania 9 first set of explanatory variables.2 The predictions of poverty derived from the models are compared with the results on poverty produced by the LSMS at a level at which the LSMS is representative. Table 1 reports the levels of poverty revealed by the LSMS, the 95 percent confidence interval (the range within which we can say with great certainty that the true poverty headcount— the poverty rate—lies), and the poverty rate predicted using census data. The poverty rate predicted using the census is within the 95 percent confidence interval of the LSMS survey in the central, coastal, and Tirana regions or domains. In the mountain domain, the census estimate is notably higher than the LSMS result. Nonetheless, the ranking of the mountain region—less well off—is consistent with the lower mean and median per capita consumption revealed through the LSMS in the mountain region relative to the other regions. A similar ranking, with the mountain areas showing lower means, is also revealed in the consumption data of the 2014 Household Budget Survey and in the gross domestic product per capita series.3 Table 1: Poverty Rates Measured in the LSMS and Estimated in the Census, by Domain, 2011–12 (percent) Domain LSMS LSMS 95 percent confidence interval Census estimate Central 12.6 10.1 15.0 13.1 Coastal 17.7 14.7 20.8 15.5 Mountain 15.1 10.9 19.4 20.6 Tirana 12.1 6.7 17.5 11.7 National 14.3 12.5 16.1 14.3 Standard errors and the level of aggregation Since June 2015, because of territorial reform, the country has been divided into 12 counties (prefecture, the first-level local administrative units), 61 municipalities (municipality, the second-level local administrative units), 72 cities and 2,980 villages/towns (villages, towns, the third-level local administrative units). Spatially, at the time of the data collected used for the mapping exercises of 2002 and 2012, Albania was divided into 12 prefectures (first-level local administrative units), 373 municipalities (second-level local administrative units), and cities, villages(together, 2. Random forest is a prediction algorithm that selects variables and predicts consumption. The method is known to produce more robust predictions. In our application, the method relies on 500 different models, and the variables that are consistently included are considered more robust predictors. See Sohnesen, Stender, and Hill (2016) for an evaluation of the method for poverty prediction. 3. A region or domain can easily show low mean per capita consumption and a low poverty rate if the levels of inequality are low. In other words, the sampled inhabitants in the region show low, but similar consumption levels even if this does not put many of them below the poverty line. This seems to be the case In the LSMS. The result is driven especially by Dibër Prefecture, which has a poverty rate that is lower than expected, which is explained by low inequality (annex C, figures C.1 and C.2). The poverty maps, in which the predictions are based on census data, do not replicate these low levels of poverty and inequality in Dibër Prefecture and estimate higher poverty rates in the mountain region or domain relative to the LSMS observations. 10 Portraits of Poverty and Inequality in Albania the third-level local administrative units). The municipalities were of two types, either urban or rural. The rural municipalities were also known as communes. Because the 2001 census and the 2011 census were both carried out based around the older system, including the 36 districts that were a second-level local administrative unit before 2000, the core results of the poverty mapping exercises for 2002 and 2012 reflect this system. Nevertheless, to produce results that are more relevant for policymaking today, mapping results for the 61 reformed second-level municipalities have also been derived by aggregating the old 373 municipalities. 4 In general, poverty can be estimated for each of the spatial levels in the 2011 census. However, the census estimates are associated with uncertainties. These are captured in standard errors. As highlighted by Tarozzi and Deaton (2008), the estimated standard errors are only correct if there is a minimal amount of spatial correlation above the cluster level. In all four (regional or domain) consumption models, the variation at the cluster level is minimal (see annex C, table C.2). Nonetheless, because poverty estimates become more specific in smaller local areas, the standard errors increase as the population of the estimate area decreases. This is illustrated in figure 1, where the standard errors in the poverty estimates are shown to decline as the number of households in the area increases. The average standard errors in the poverty estimates for municipalities and communes are 0.04. In comparison, the LSMS domain (regional) standard errors are in the 0.01 to 0.03 range. As expected, the municipal poverty estimates are thus associated with greater standard errors relative to the domain estimates based on the LSMS data, particularly in those areas with fewer than 3,000 inhabitants (log 8 in figure 1). Among the old municipalities, 130 have fewer than 3,000 inhabitants each, while only one of the new municipalities and none of the old districts have fewer than 3,000 inhabitants. Substantial standard errors are therefore associated with the poverty estimates for a number of communes and municipalities. Figure 1: Standard Errors, Municipality and Commune Estimates 0.2 SE of poverty headcount 0.15 0.1 0.05 0 5 6 7 8 9 10 11 12 13 14 log of population in location Note: The figure is based on calculations of standard errors in the poverty rates for communes and municipalities. The figure shows results for both old and new municipalities. 4. The aggregation may not correspond precisely to the current administrative divisions because the old second-level districts are now defunct, and the former rural municipalities or communes have been abolished and are now counted as third-level units, that is, neighborhoods (lagje) or villages (fshat), within the new municipalities. These new results may nonetheless be more relevant for policy making today because they more closely mirror the administrative units on which policy making and poverty initiatives will become focused henceforward. Subsection 4.2 provides some of these new results. Portraits of Poverty and Inequality in Albania 11 CHAPTER 3 12 Portraits of Poverty and Inequality in Albania Results This section summarizes the main results of the poverty mapping exercise in Albania. It presents the main maps produced on the poverty rate (headcount), the number of the poor individuals, the average monthly per capita consumption, and the Gini coefficient (inequality). The maps on poverty in the prefectures, along with descriptive material on the results are presented in the Main Annex (available online). 3.1 General analysis of the results of the mapping exercise Prefectures The poverty rate (headcount) in Albania was estimated at 14.3 percent in 2012, the same to the national poverty rate estimated through the LSMS 2012. The highest poverty rate was in Kukës Prefecture (around 22.0 percent), and the lowest rate was in Gjirokastër Prefecture (around 8.0 percent). There were 398,131 poor individuals in the country. The number was higher in the central region (153,968 poor individuals), and the lowest number was in the mountain region (53,337 poor individuals). Tirana Prefecture had the highest number of poor people (94,101), Gjirokastër Prefecture had the lowest number (5,988). Average per capita monthly consumption in the country in 2012 was 8,477 ALL. Prefectures with the highest level of consumption were Gjirokastër (10,190 ALL), Korçë (9,260 ALL), and Berat (8,785 ALL), while the prefectures with the lowest level of consumption were Kukës (7,126 ALL), Dibër (7,551 ALL) and Elbasan (8,192 ALL). Municipalities and communes Poverty rates varied across communes, from 2.6 percent in the commune of Zagori in Gjirokastër Prefecture to 38.5 percent in the commune of Kalis in Kukës Prefecture (map 2). The highest poverty rates were in the communes and districts in the northeast of the country, where the darker red color is more expansive in map 2. In the south and southeast of the country, poverty rates were substantially lower. The poverty rates were higher in the communes and districts of the mountain region (20.6 percent), and the lowest rates were in the Tirana region (11.7 percent). The prefectures of Durrës, Kukës, and Tirana showed large differences in the poverty rates across municipalities and communes. In the prefecture of Durrës, Portraits of Poverty and Inequality in Albania 13 Map 2: Poverty Maps, Communes and Districts, Albania, 2012 a. Poverty rates (headcount), communes b. Poverty rates (headcount), districts Poverty Head Count: Poverty Head Count by Districts MALËSI E MADHE TROPOJË HAS SHKODËR PUKË KUKËS Poverty Head Count: Poverty Head Count LEZHË by Districts MIRDITË DIBËR KURBIN MAT MALËSI E MADHE TROPOJË KRUJË BULQIZË DURRËS HAS TIRANË SHKODËR PUKË KUKËS LIBRAZHD KAVAJË PEQIN ELBASAN LEZHË MIRDITË LUSHNJE POGRADEC DIBËR KUÇOVË GRAMSH KURBIN MAT FIER BERAT KRUJË KORÇË BULQIZË DEVOLL DURRËS MALLAKASTËR SKRAPAR TIRANË VLORË TEPELENË Head Count ratio (%) KAVAJË LIBRAZHD Head Count ratio (%). KOLONJË PËRMET PEQIN < 8.0 < 8.0 ELBASAN 8.0 - 12.0 8.0 - 12.0 LUSHNJE 12.1 - 16.0 GJIROKASTËR POGRADEC 16.1 - 20.0 12.0 - 16.0 KUÇOVË GRAMSH 20.1 - 24.0 DELVINË 18.0 - 20.0 FIER 24.1 - 28.0 BERAT > 20.0 KORÇË > 28.0 DEVOLL SARANDË Prefecture Boundary MALLAKASTËR SKRAPAR Prefecture Boundary Commune/Municipality Boundary District Boundary Source: Head Count ratio VLORË(%). TEPELENË Head Count ratio (%) The boundaries of communes and municipalities KOLONJË Population and Housing Census 2011 PËRMET Source: have been designed for statistical purpose and The boundaries of communes and municip Living Standard Measurement Survey – LSMS 2012 < 8.0 andnot Populationmay reflect Census Housing 2011 exactly the territory of the local units. < 8.0 have been designed for statistical purpose 8.0 - 12.0 Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the lo 12.1 - 16.0 8.0 - 12.0 GJIROKASTËR 16.1 - 20.0 12.0 - 16.0 20.1 - 24.0 DELVINË 18.0 - 20.0 24.1 - 28.0 > 20.0 > 28.0 Prefecture Boundary SARANDË Prefecture Boundary Commune/Municipality Boundary District Boundary Source: Population Source: and Housing The boundaries Census of communes and2011 municipalities The boundaries of communes and municipalities have been designed for have been designed for statistical purpose and – LSMS 2012 Living Standard Measurement Population and Housing may Census not reflect Survey 2011 exactly – LSMS the territory 2012 of the local units. notboundaries statistical purpose and may The reflect of communes and municipalities exactly the territory of the local unit have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 14 Portraits of Poverty and Inequality in Albania poverty rates ranged from 9.0 percent in the commune of Bubq to 21.0 percent in the municipality of Sukth. In the prefecture of Kukës, the lowest poverty rate was in Bajram Curri Municipality (10.0 percent), and the highest rate was in Kalis Commune (38.5 percent), which is also the poorest commune in the country. In the prefecture of Tirana, poverty rates varied from 9.2 percent in the municipality of Tirana to 25.2 percent in the municipality of Kamëz. The prefectures with the largest gap between the lowest and highest poverty rates were Dibër, Elbasan, and Kukës. In the prefecture of Kukës, the lowest poverty rate was in Bajram Curri Municipality (10.0 percent), and the highest rate was in Kalis Commune (38.5 percent), the poorest commune in the country. The lowest poverty rate in Elbasan Prefecture was in Librazhd Municipality (8.7 percent), and the highest was in Orenjë Commune (30.4 percent). In Dibër Prefecture, the lowest poverty rate was in the municipality of Burrel (11.2 percent), and the highest rate was in Sllovë Commune (29.4 percent). The prefectures of Berat, Fier, and Gjirokastër exhibited the narrowest gaps between the lowest and the highest poverty rates, less than 12 percentage points. In Gjirokastër Prefecture, the poverty rates ranged from 2.6 percent in Zagori Commune to 14.3 percent in the commune of Krahës. In Fier Prefecture, the lowest poverty rate occurred in the municipality of Fier (11.0 percent), and the highest rate was in Selitë Commune (22.1 percent). Durrës, Elbasan, Kamëz, and Tirana were among the municipalities with the higher number of poor people, as well as larger shares of the country’s population so that, relative to the country average, the number of poor individuals was higher in these municipalities. The number of the poor was low in the communes of Odrie, Pogon and Zagori in Gjirokastër Prefecture. The lowest monthly per capita consumption in 2012 was in the communes of Kalis (5,839 ALL), Ujmisht (5,983 ALL), and Surroj (6,124 ALL) in Kukës Prefecture. The highest values were in the communes of Zagori (13,216 ALL), Pogon (11,936 ALL) and Odrie (11,839 ALL) in Gjirokastër Prefecture. The level of inequality measured using the Gini coefficient ranged from 19.7 percent in the commune of Kalis, which is also the poorest commune in the country, to 30.0 percent in the commune of Farkë. The communes of Grykë-Çajë, Kala e Dodës, and Ujmisht recorded the lowest Gini coefficients, around 20 percent. The commune of Dajt and the municipalities of Durrës, Lezhë, Sarandë, and Tirana had the highest Gini coefficients in the country, around 28 percent. Portraits of Poverty and Inequality in Albania 15 CHAPTER 4 16 Portraits of Poverty and Inequality in Albania Using poverty maps in policy making Understanding which areas have higher poverty rates can potentially allow the more accurate and efficient targeting of resources for poverty reduction. For this purpose, it is important to understand why these areas are relatively poorer and address the associated issues. The reasons are likely to vary from place to place and may include inadequate infrastructure, lack of economic activity, and an insufficiently skilled workforce. The results on poverty shown in the poverty maps can be easily mapped and overlaid, for example, with spatial data on roads and topographical elevations, health and educational facilities, population density, and so on to help reveal links between poverty and welfare indicators. Likewise, the maps may force more thinking in subnational and national decision making and policy making on how best to balance the targeting of poor areas and poor people to combat poverty and social exclusion, but also improve standards of living and help guide the allocation of government expenditures in other areas of development. Led by cutting-edge research and widespread empirical application of the poverty mapping methodology, more than 60 countries have, since the late 1990s, gained experience in the small area estimation of poverty illustrated by poverty maps. Spotlight 1 presents examples of applications in other countries to illustrate innovative ways to use the information provided by the maps in policy making. The next paragraphs provide initial insights into using the poverty maps for policy making in Albania, but hopefully this is only the beginning as the findings of this report are actively employed in policy making in Albania. 4.1 The poverty maps for policy in Albania We illustrate how the results of the poverty mapping exercise in Albania can be useful for policy making in two ways. First, we present the poverty maps revised according to the new local administrative units established through the territorial reform launched in June 2015 to demonstrate that poverty maps can be adapted and made useful within an altered geographical area administrative structure. Second, we showcase how census or survey data on other areas of policy interest (education coverage, road infrastructure, public welfare expenditure, and so on) may be overlaid on the maps to help perform a triage among small geographical areas according to particular characteristics to target benefits, implement antipoverty programs, support appropriate infrastructure development, and so on. Portraits of Poverty and Inequality in Albania 17 Poverty maps of the new local administrative units The new territorial division of Albania divide the country into 61 municipalities. The highest level of poverty is in the municipality Kamëz (about 24 percent), followed by municipality of Has (23.3 percent) and the municipality of Kukës (23.2 percent). The lowest level of poverty is recorded in the municipalities of Pustec (5 percent), Libohovë (6.7 percent) and Gjirokastra (6.8 percent). Municipalities with the largest number of poor are municipalities of Kamëz (24,695), Durres (27,773 ) and the municipality of Tirana (55,561). The lowest number of poor recorded in municipalities of Pustec (158), Libohovë (246) and Dropull (273). The lowest level of the average monthly per capita consumption were in municipalities of Kukës (7,067 ALL), Has (7,192 ALL) and Dibër (7,232 ALL). While the highest monthly per capita consumption were in the municipalities of Libohovë (10,895 ALL), Pustec (11,213 ALL) and Gjirokastër (11,215 ALL). The municipalities with the lowest level of inequality (Gini coefficent) were Bulqizë, Kukës and Dibra, lower than 8 percent, and the highest level in the municipalities of Vlorë, Durrës and Sarandë, around 27 percent. 18 Portraits of Poverty and Inequality in Albania Map 3: Poverty Rate and Inequality, New Municipalities, 2012 Number of poor b. Number of the poor Poverty Head a. Poverty Count rate (headcount) Poverty Head Count Number of poor individuals Persons < 60 600 - 150 1501 - 300 3001 - 500 Head Count ratio (%). 5001 - 900 < 8.0 9001 - 1300 8.0 - 11.0 13001 - 3000 11.1 - 15.0 15.1 - 20.0 > 3000 > 20.0 Prefecture Boun Prefecture Boundary Municipality Bou Municipality Boundary Persons Head Count ratio (%). Source: < 600 Source: Population and Housing Census of The boundaries 2011municipalities have been The boundaries of munici < 8.0 600 - 1500 Population and Housing Census 2011 Living Standard designed Survey –purpose for statistical Measurement LSMS 2012and for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. exactly the territory of the 8.0 - 11.0 1501 - 3000 11.1 - 15.0 3001 - 5000 15.1 - 20.0 5001 - 9000 > 20.0 9001 - 13000 Prefecture Boundary 13001 - 30000 Municipality Boundary > 30000 The boundaries of municipalities have been designed for statistical purpose and S 2012 may not reflect exactly the territory of the local units. Prefecture Boundary Municipality Boundary Portraits of Poverty and Inequality in Albania Source: 19 Population and Housing Census 2011 The boundaries of municipalities have been designed Living Standard Measurement Survey – LSMS 2012 for statistical purpose andmay not reflect exactly the territory of the local units. c. The average monthly per capita consumption d. Gini coefficient (inequality) Per Capita Consumption Gini Coefficient Gini Coefficient Per Capita Consumption Per capita consumption ( ALL/month ) < 7600 7600 - 8500 Gini index (%) 8501 - 9300 < 10.0 9301 - 10500 10.0 - 19.0 > 10500 19.1 - 24.0 Prefecture Boundary 24.1 - 26.0 > 26.0 Municipality Boundary Prefecture Boundary Gini coefficient is a measure of inequality with values between 0 and 100. Municipality Boundary Source: The boundaries of municipalitieshave been designed Population and Housing Census 2011 Source: The boundaries of communes and municipali Per capita consumption ( ALL/month ) for statistical purpose Giniandmay reflect exactly the not(%) index Living Standard Measurement Survey – LSMS 2012 territory2011 Population and Housing Census of the local units. have been designed for statistical purpose an Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the loca < 7600 < 10.0 7600 - 8500 10.0 - 19.0 8501 - 9300 19.1 - 24.0 9301 - 10500 24.1 - 26.0 > 26.0 > 10500 Prefecture Boundary Prefecture Boundary Gini coefficient is a measure of inequality with values between 0 and 100. Municipality Boundary Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: Population and Housing The boundaries Census 2011 of municipalitieshave been designed The boundaries of communes and municipalities have been designed for Measurement Living Standard for statistical purposeSurvey andmay not – LSMS 2012 reflect exactly the statistical purpose and may not reflect exactly the territory of the local unit territory of the local units. 20 Portraits of Poverty and Inequality in Albania Overlaying poverty maps with other data of policy interest The final maps presented in this section illustrate how data from other sources and on other topics of interest may be placed on top of the maps to enrich policymakers’ evidence base on policy relevant challenges to reduce poverty and inequality. Map 4, for instance, combines poverty headcount data with Census information on the percentage of households that receive social assistance in each geographical area. It shows that the percentage of households with social assistance as a source of income is higher in poorer areas. Map 5 combines information on poverty headcounts and number of people with low educational attainment5, while Map 6 shows that there is not a unique pattern across Albania in terms of poverty and sector of employment concentration of the population. Source of income of households: Source of income of househ Social assistance Social assistance Map 4: Poverty Rate and Percentage of Households with Social Assistance, 2012 Households with social assistance in % Households with social assistanc < 7.0 < 7.0 7.0 - 12.0 7.0 - 12.0 12.1 - 20.0 12.1 - 20.0 20.1 - 35.0 20.1 - 35.0 > 35.0 > 35.0 Poverty Headcount in % Poverty Headcount in % < 8.0 < 8.0 8.0 - 11.0 8.0 - 11.0 11.1 - 15.0 11.1 - 15.0 15.1 - 20.0 15.1 - 20.0 > 20.0 > 20.0 Prefecture Boundary Prefecture Boundary Municipality Boundary Municipality Boundary 5. For the The education boundaries map, have of municipalities results been must Source:be interpreted with caution as the level of education is used as part of The boundaries of municipalities have be Census 2011 designed for statistical purpose and Population and Housing Census 2011 designed for statistical purpose and ment Survey - LSMS 2012 the model may generating the not reflect exactly thepoverty head territory of the counts, localLiving hence, units Standard by construction these two levels would be correlated. Measurement Survey - LSMS 2012 may not reflect exactly the territory of the Portraits of Poverty and Inequality in Albania 21 Map 5: Poverty Rate and Number of Map 6: Poverty Rate and People with Primary and Lower Share of People Employed by Secondary Education, 2012 Educational attainment: Economic Sector, 2012 Employment by econom Primary and lower secondary Employment by economic activities Educational attainment: Primary and lower secondary Employment by economic activities Number of persons Educational attainment: <= 10000 Primary and lower secondary 10001 - 20000 20001 - 35000 Employment in % 35001 - 70000 Industry > 70000 Services Agriculture Number of persons Poverty Headcount in % Poverty Headcount in % <= 10000 < 8.0 < 8.0 10001 - 20000 8.0 - 11.0 8.0 - 11.0 11.1 - 15.0 20001 - 35000 11.1 - 15.0 15.1 - 20.0 15.1 - 20.0 Employment in % > 20.0 35001 - 70000 > 20.0 Prefecture Bo Prefecture Boundary Municipality B Number of persons Industry Municipality Boundary > 70000 Services <= 10000 Agriculture Source: Source: The boundaries of municipalities have been The boundaries of municipalit Population and Housing10001 Census 2011 - 20000 Population and Housing designed forCensus 2011 statistical purpose and designed for statistical purpos Poverty Headcount in % Living Standard Employment in % 2012 Poverty Headcount in % Living Standard Measurement Survey - LSMS 2012 mayMeasurement not reflect Survey exactly LSMS the-territory of the local units may not reflect exactly the ter < 8.0 20001 - 35000 < 8.0 8.0 - 11.0 8.0 - 11.0 Industry 11.1 - 15.0 11.1 - 15.0 35001 - 70000 Services 15.1 - 20.0 15.1 - 20.0 Agriculture > 20.0 > 20.0 Prefecture Boundary Poverty Headcount in % Prefecture Boundary > 70000 Municipality Boundary < 8.0 Municipality Boundary 8.0 - 11.0 Poverty Population Source: Headcount in and % Housing Census 2011 The boundaries of municipalities have been The boundaries of communes11.1 and municipalities have been designed for - 15.0 Housing Census 2011 designed for statistical purpose and Measurement Survey - LSMS 2012 < 8.0 Source: Survey – LSMS 2012 Living Standard Measurement may not reflect exactly the territory of the localstatistical units purpose and may not reflect 15.1 - 20.0 exactlyThe the territory boundaries ofof the local municipalities unit have been Population and Housing Census 2011 designed for statistical purpose and 8.0 - 11.0 Living Standard Measurement Survey - LSMS 2012 > 20.0 may not reflect exactly the territory of the local units Prefecture Boundary 11.1 - 15.0 22 Municipality and Portraits of Poverty Inequality in Albania Boundary 15.1 - 20.0 > 20.0 Source: Prefecture Boundary The boundaries of municipalities have been Spotlight 1. Poverty maps and policy making around the world Morocco The Moroccan government requested the support of the World Bank in 2002 to learn how to use poverty mapping techniques and produce a detailed, disaggregated map of poverty throughout the country.6 At the time, the Bank was scheduled to undertake a poverty report on the country. This enabled the emergence of an active dialogue on how to improve targeting in social spending. When the first poverty map and other analyses were completed, key policy makers had already been sensitized to the significance of poverty maps. The need for finer geographical targeting of the poor was intensified because of two factors. First, because of the threat of terrorism, renewed attention was being focused on the vulnerability of the urban poor and the phenomenon of rural-urban migration. More information was needed on pockets of poverty and vulnerability. Second, because of a growing budget deficit, there was strong pressure to make public expenditures more effective. The desire to minimize benefit leakage to the nonpoor had gained strength. The poverty mapping program proceeded for almost a year on technical and policy levels. The technical focus permitted a transfer of the capacity to construct poverty maps to local experts, and the policy focus stimulated interest in the potential utility of the maps for policy purposes; if the maps were to have an impact on policy, they had to appeal to decision makers who would be able to apply them to realize the country’s policy goals. The World Bank published a poverty report two months after the government’s poverty map report (Morocco, Planning Commission 2004; World Bank 2004). Both publications provided analysis of the spatial aspects of poverty and inequality, but also other issues. Eventually, the government constructed a second, updated poverty map. That poverty varied greatly not only among provinces, but also within provinces was an eye-opener for policy makers and led to a reappraisal of targeting strategies. Whereas the major poor-area development project . . . had relied on province-level targeting, the new information on poor communes provided the government with fresh insights into the possible design of a more finely targeted program, the National Initiative for Human Development. (Litvack 2007, 216) The National Initiative for Human Development was launched in May 2005. The government announced that $1 billion would be allocated to the program, half of which would go to efforts to target extra resources to the poorest 360 rural communes and poorest 250 urban neighborhoods. The maps were to be used for targeting purposes. The maps also began to play an interesting role in promoting local governance. Several governors had been surprised by some of the poverty map data and had undertaken site visits to confirm the data. The poverty maps thus contributed to a much broader agenda of transparency and good governance. 6. This brief case study on Morocco owes much to World Bank (2015). Portraits of Poverty and Inequality in Albania 23 European Union, 2012–2014 In 2012–14, the World Bank assisted in the creation of poverty maps of Estonia, Hungary, Latvia, Poland, Romania, the Slovak Republic, and Slovenia as part of an ongoing European Commission effort to produce poverty maps of all countries in the European Union. The maps portray subnational geographical areas, such as municipalities, counties, and districts, and highlight the small areas most likely to exhibit the highest risk of poverty in each of these member states. The new maps complement other information on the correlates of poverty that enhances regional policy and program design. They have effectively guided decision making and policy making at the subnational and national levels in member states of the European Union and Map: Poverty Mapping in the European Union have assisted the European Commission in allocating Source: Simler and van der Weide 2015. national and regional development and financing program funds efficiently to areas with the greatest need. As the national partners of the project, government authorities, mainly experts at national statistical institutes and ministries, have benefited from substantial knowledge and skill spillovers about poverty mapping through the project. The spillovers have gained prominence through academic research by catalyzing greater communication and collaboration. Other examples The following is an alphabetical list of selected countries in which poverty mapping has been used to achieve a representative policy end. The name of the country is followed by the year of the map and a brief description of the policy applications. Bolivia, 2002–2003 In Bolivia, the method was used to create consumption poverty maps at the level of municipalities (Arias and Robles 2007). The purpose was to generate local indicators of monetary poverty and consumption inequality for the measurement of municipal disparities. On-site interviews with potential map users in 2006 revealed that the maps were being used in planning. Thus, national and international institutions were using the maps to plan strategies, activities, and programs. The latter included the Swiss Agency for Development and Cooperation and the United Nations Development Programme. The maps were also being used to target public productive and social investments and regional development resources, including a community investment program funded by the Inter-American Development Bank. Bulgaria, 2003, 2005 The first poverty and inequality maps on Bulgaria measured and illustrated mean per capita consumption by municipality, the poverty headcount index, and the poverty depth index (Gotcheva 2007). Among the many initiatives that relied on the maps were targeted antipoverty interventions for the poorest municipalities, and the mainstreaming of antipoverty policies directed at the municipalities identified as the poorest into national strategic documents aimed at the reduction of poverty, the promotion of employment, and the elimination of social exclusion. A targeted training and employment program created well over 1,000 jobs in poor municipalities. Cambodia, 2002 The primary objective of the poverty mapping exercise in Cambodia was, from the outset, to develop a 24 Portraits of Poverty and Inequality in Albania tool for policy making and, especially, for the allocation of resources (Fujii 2007). Interviews conducted among potential users in 2006 revealed that, among the many applications, the maps were being used by the World Food Programme to create overlay maps on infrastructure, flood and draught vulnerability, and education to link poverty and these issues that highlighted small geographical areas with particular needs. The resulting nutrition maps and other sorts of maps, along with field observations, became a fundamental tool of the World Food Programme in the identification of areas for program intervention. Yunnan Province, China, 2002–2003 Poverty mapping was undertaken in Yunnan Province—selected because of its ethnic diversity and relatively high poverty incidence—to test whether the methodology might produce reliable poverty incidence estimates down to the township level in rural areas and down to the district level in urban areas (Ahmad and Goh 2007a). The National Development and Reform Commission used the map to review the county-level allocation of project funding, as well as the incidence of poverty in all counties. The World Bank and the U.K. Department for International Development have used the poverty maps to help select beneficiary areas for a community development project. The data issues uncovered during the exercise were particularly valuable in improving data collection at the National Bureau of Statistics. Ecuador, 1990, 2001 A central goal in constructing the 2001 poverty maps was to enable users to track changes in small area estimates of poverty since the creation of the maps in 1990 (Araujo 2007). In addition, Ecuador independently produced poverty maps on unmet basic needs and poverty maps on poverty, health, education, employment, and the environment. The maps were applied on a small scale to determine the appropriate location for projects in employment, microfinance, and small- and medium-enterprise development; urban reconstruction projects in historic neighborhoods; the construction of maps on food security and food vulnerability; geographical targeting and donor projects; targeting the allocation of educational inputs; determining the location of early childhood development interventions; internal planning and targeting investments at ministries; and targeting emergency social investments. Indonesia, 2002–2004 The national poverty maps have been used in various applications by government agencies, donors, and nongovernmental organizations (Ahmad and Goh 2007b). These applications include facilitating budget estimates of unconditional cash transfers to the poor, the selection of program beneficiary areas, cross- checking and referencing government poverty data, the production of follow-up nutrition maps, identifying poor districts for local development planning, and guiding policy advice to the government in donor projects. Sri Lanka, 2005 The poverty maps produced in Sri Lanka have not only enhanced public awareness of the significant regional inequalities in the country, but also encouraged policy makers to take appropriate steps to address the severe deprivation that still prevails in rural and remote areas (Vishwanath and Yoshida 2007). They are having an effect on policies, lending strategies, and research. In particular, the poverty maps are influencing road projects, helping revive the government’s initiative to improve targeting in benefit transfer programs, and identifying the poorest areas to raise benefits there. Vietnam, 1997, 1999, 2003, 2005 Interviews were conducted in 2006 to determine impacts of the various poverty mapping exercises in Vietnam (Swinkels and Turk 2007). These impacts included improvements in understanding the nature of poverty in the country, better targeting of resources, and facilitating reviews of poverty measurement techniques. Portraits of Poverty and Inequality in Albania 25 CHAPTER 5 26 Portraits of Poverty and Inequality in Albania Conclusions This report presents results for the new poverty maps in Albania using 2012 LSMS and 2011 census data and incorporating improvements to the poverty mapping technique. A poverty map is a geographical profile that shows where poverty is concentrated in a country by providing information on lower geographic levels, in the case of Albania, in the municipalities and communes. The analysis of maps is done based on four main indicators: poverty rate, number of poor people, average monthly per capita consumption and inequality measured by the Gini coefficient. These four indicators are presented on country, region, 12 prefectures level and for each municipality / commune. Poverty maps may be applied in policy making in three key ways: (1) as a benchmark against existing resource allocation criteria, for example, whether the allocation of social-assistance block grants according to previously established criteria correlate with an appropriate allocation based on current poverty rates; (2) as a tool in targeting public spending; and (3) for the provision of data to monitor the progress toward achieving particular government welfare goals. Nongovernmental organizations and international bilateral and multilateral institutions can also rely on poverty maps in supplying advisory services to local governments and donor agencies and in designing joint intervention strategies. Summary of main results In Albania, 14.3 percent (2012) of population lived in poverty. The highest level of poverty was concentrated in the northeast of the country, while in the south and southeast this indicator is much lower. The level of poverty was highest in Mountain region (20.6 percent) and the lowest level was in Tirana region (11.7 percent). The highest level of poverty was in the prefecture of Kukës (around 22 percent) and the lowest level was recorded in the prefecture of Gjirokastër (around 8 percent). Poverty level varied from 2.6 percent, in the commune of Zagori in the prefecture of Gjirokastër, to 38.5 percent in the commune of Kalis in the prefecture of Kukës. Looking to the fluctuations of poverty level inside each prefecture it was noted that prefectures of Kukës, Tiranë and Durrës had big differences of poverty level between municipalities/communes. According to 2012 data, the number of poor people in the country was 398.131 individuals. The number of poor people was higher in the Central region (153.968 poor individuals) and the lowest number was in the Mountain region (53.337 poor individuals). The prefecture with the highest number of poor people was Tirana Portraits of Poverty and Inequality in Albania 27 (94.101 poor individuals) while the prefecture of Gjirokastër had the lowest number of the poor people (5.988 poor individuals). The number of poor people was generally concentrated on those municipalities/ communes where the population density was higher. The highest number of poor people was in the prefecture of Tirana, in the municipality of Tirana with 37.672 poor individuals and with a level of poverty 9.2 percent, followed by municipalities of Kamëz and Durrës with more than 15.000 poor individuals. The number of poor people was lower in communes Zagori, Pogon and Odrie, part of Gjirokastër prefecture. According to 2012 data, the average monthly per capita consumption in the country was 8.477 ALL. The average per capita consumption level seemed higher in the south of the country compared with other areas. The highest value of the average per capita consumption was in the Central region (8.991 ALL) while the lowest value was recorded in the Mountain region (7.368 ALL). Prefectures with the highest level of per capita consumption were Gjirokastër (10.190 ALL), Korça (9.260 ALL) and Berat (8.785 ALL), while the prefectures with the minimum level of per capita consumption were those of Kukës (7.126 ALL), Diber (7.551 ALL) and Elbasan (7.979 ALL). The lowest values of per capita consumption were recorded in the communes of Kalis (5.839 ALL), Ujmisht (5.983 ALL) and Surroj (6.124 ALL) of Kukës prefecture. The highest values were recorded in the communes of Zagoria (13.216 ALL), Pogon (11.936 ALL) and Odrie (11.839 ALL) of Gjirokastër prefecture. The level of inequality measured by Gini coefficient was around 24 percent. The highest level of inequality was in the Tirana region (25.7 percent) and the lowest level was in the Mountain region (22.7 percent). The highest level of inequality was recorded in the prefecture of Shkodër (25.1 percent) and the lowest level in the prefecture of Kukës (22.4 percent). The level of inequality ranged from 19.7 percent in the commune of Kalis, which was also the poorest commune of the country, to 30 percent in the commune of Farka. The communes of Gryke Caj, Ujemisht and Kala e Dodës also recorded lowest level of Gini Coefficient, around 20 percent. After commune of Farka, Municipalities of Lezha, Durrës, Saranda, Tirana and the commune of Dajti had the highest levels of inequality in the country, around 28 percent. Inequality is in the same direction with per capita consumption and in the opposite direction with the level of poverty. However, if we take into consideration a detailed level of geographical disaggregation we note that this condition is not always fulfilled. 28 Portraits of Poverty and Inequality in Albania Portraits of Poverty and Inequality in Albania 29 References Ahmad, Yusuf, and Chor-ching Goh. 2007a. “Poverty Maps of Yunnan Province: Uses and Lessons for Scaling-Up in China.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 143–52. Washington, DC: World Bank. ———. 2007b. “Indonesia’s Poverty Maps: Impacts and Lessons.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 177–87. Washington, DC: World Bank. Araujo, María Caridad. 2007. “The 1990 and 2001 Ecuador Poverty Maps.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 153–76. Washington, DC: World Bank. Arias, Omar, and Marcos Robles. 2007. “The Geography of Monetary Poverty in Bolivia: The Lessons of Poverty Maps.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 67–89. Washington, DC: World Bank. Betti Gianni, Andrew Dabalen, Celine Ferrè, and Laura Neri. 2010. “Further Updating Poverty Mapping in Albania.” Paper presented at the World Bank conference, “Poverty and Social Inclusion in the Western Balkans,” Brussels, December 14–15. Betti, Gianni, Laura Neri, and Francesca Ballini. 2003. “Poverty and Inequality Mapping in Albania: Final Report.” Unpublished working paper, World Bank and Institute of Statistics, Tirana, Albania. Dabalen, Andrew, and Celine Ferrè. 2008. “Updating Poverty Maps: A Case Study of Albania.” Unpublished working paper, World Bank, Washington, DC. EC (European Commission). 2010. “Europe 2020: A Strategy for Smart, Sustainable, and Inclusive Growth.” Communication from the Commission (March 3), EC, Brussels. http://eur-lex.europa.eu/ LexUriServ/LexUriServ.do?uri=COM:2010:2020:FIN:EN:PDF. Elbers, Chris, Jean O. Lanjouw, and Peter F. Lanjouw. 2002. “Micro-Level Estimation of Welfare.” Policy Research Working Paper 2911, World Bank, Washington, DC. ———. 2003. “Micro-Level Estimation of Poverty and Inequality.” Econometrica 71 (1): 355–64. Elbers, Chris, and Roy van der Weide. 2014. “Estimation of Normal Mixtures in a Nested Error Model with an Application to Small Area Estimation of Poverty and Inequality.” Policy Research Working Paper 6962, World Bank, Washington, DC. Foster, James, Joel Greer, and Erik Thorbecke. 1984. “A Class of Decomposable Poverty Measures.” Econometrica 52 (3): 761–66. Fujii, Tomoki. 2007. “To Use or Not to Use? Poverty Mapping in Cambodia.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 125–42. Washington, DC: World Bank. Gotcheva, Boryana. 2007. “The Poverty Mapping Exercise in Bulgaria.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 90–124. Washington, DC: World Bank. Hungary, Ministry for National Economy. 2015. “National Reform Programme 2015 of Hungary.” Ministry for National Economy, Budapest. INSTAT (Albanian Institute of Statistics). 2004. “Poverty and Inequality Mapping in Albania.” Report, INSTAT, Tirana, Albania. ———. 2012. Population and Housing Census, 2011: Main Results. Tirana, Albania: INSTAT. 30 Portraits of Poverty and Inequality in Albania ———. 2013. “Albania: Trends in Poverty, 2002-2005-2008-2012.” September, INSTAT, Tirana, Albania. ———. 2014. “Albania, Dwelling and Living Conditions.” May, INSTAT, Tirana, Albania. INSTAT (Albanian Institute of Statistics) and IOM (International Organization for Migration). 2014. “Migracioni i kthimit dhe ri-integrimi në Shqipëri, 2013” (Return migration and reintegration in Albania, 2013). INSTAT and IOM, Tirana, Albania. http://www.instat.gov.al/al/themes/popullsia/ publications/books/2014/migracioni-i-kthimit-dhe-ri-integrimi-në-shqipëri-2013.aspx. Kilic, Talip, and Thomas Pave Sohnesen. 2015. “Same Question but Different Answer: Experimental Evidence on Questionnaire Design’s Impact on Poverty Measured by Proxies.” Policy Research Working Paper 7182, World Bank, Washington, DC. KSH (Hungarian Central Statistical Office) and World Bank. 2013. “Poverty Mapping in Hungary: Making Better Policies through Better-Targeted Interventions.” Poverty Mapping Brief, World Bank, Washington, DC. Lemieux, Thomas. 2002. “Decomposing Changes in Wage Distributions: A Unified Approach.” Canadian Journal of Economics 35 (4): 646–88. Litvack, Jennie. 2007. “The Poverty Mapping Application in Morocco.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 208–24. Washington, DC: World Bank. Molina, Isabel, and J. N. K. Rao. 2010. “Small Area Estimation of Poverty Indicators.” Canadian Journal of Statistics 38 (3): 369–85. Morocco, Planning Commission. 2004. “Carte de la pauvreté communale.” Report, Planning Commission, Rabat, Morocco. Simler, Kenneth, and Roy van der Weide. 2015. “Poverty Mapping: Concepts, Policy Applications, and Preliminary Results in Croatia.” Report presented during a World Bank mission, Zagreb, Croatia, December 18. Sohnesen, Thomas Pave and Niels Stender. 2016. “Is Random Forest a Superior Methodology for Predicting Poverty?: An Empirical Assessment.” World Bank Policy Research Working Paper (7612). Swinkels, Rob, and Carrie Turk. 2007. “Poverty Mapping: The Experience of Vietnam.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 261–86. Washington, DC: World Bank. Tarozzi, Alessandro, and Angus S. Deaton. 2009. “Using Census and Survey Data to Estimate Poverty and Inequality for Small Areas.” Review of Economics and Statistics 91 (4): 773–92. Vishwanath, Tara, and Nobuo Yoshida. 2007. “Poverty Maps in Sri Lanka: Policy Impacts and Lessons.” In More Than a Pretty Picture: Using Poverty Maps to Design Better Policies and Interventions, edited by Tara Bedi, Aline Coudouel, and Kenneth Simler, 225–40. Washington, DC: World Bank. World Bank. 2004. “Kingdom of Morocco Poverty Report: Strengthening Policy by Identifying the Geographic Dimension of Poverty.” Report 28223-MOR, Social and Economic Development Group, Middle East and North Africa Region, World Bank, Washington, DC. ———. 2012. “Recommendations for the Sample Design and Estimation Procedures for the 2012 Albania LSMS.” World Bank, Tirana, Albania. ———. 2015. “Pinpointing Poverty on Maps of Europe.” Draft report (July 13), World Bank, Washington, DC. Zhao, Qinghua, and Peter Lanjouw. 2009. “Using PovMap2: A User’s Guide.” World Bank, Washington, DC. Portraits of Poverty and Inequality in Albania 31 AnnexES 32 Portraits of Poverty and Inequality in Albania | Annexes Results of the 2012 Poverty Maps Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption BERAT BERAT 36,374 0.124 0.01 0.259 0.003 9,406.61 9,269.28 211.65 BERAT KUTALLI 9,643 0.145 0.02 0.240 0.006 2,314.62 8,443.02 311.52 BERAT LUMAS 3,975 0.154 0.04 0.239 0.008 949.50 8,309.24 416.36 BERAT VELABISHT 8,469 0.135 0.02 0.235 0.005 1,992.95 8,495.93 312.29 BERAT OTLLAK 9,203 0.121 0.02 0.245 0.005 2,250.88 8,926.60 338.23 BERAT POSHNJË 7,366 0.123 0.02 0.240 0.005 1,767.89 8,809.21 326.61 BERAT ROSHNIK 2,513 0.123 0.04 0.238 0.009 597.89 8,786.14 595.11 BERAT SINJË 3,365 0.178 0.04 0.233 0.008 783.75 7,872.54 406.04 BERAT TERPAN 1,699 0.160 0.04 0.224 0.009 380.86 7,900.38 467.67 BERAT URA VAJGURORE 7,232 0.126 0.03 0.245 0.006 1,774.81 8,908.28 404.98 BERAT VERTOP 4,914 0.124 0.03 0.238 0.006 1,167.30 8,707.51 419.57 BERAT CUKALAT 3,045 0.151 0.03 0.227 0.008 690.21 8,139.50 442.85 BULQIZË BULQIZË 8,171 0.136 0.03 0.226 0.006 1,847.12 8,204.99 379.96 BULQIZË FUSHË-BULQIZË 3,342 0.244 0.02 0.231 0.005 770.35 7,110.53 320.65 BULQIZË GJORICË 4,213 0.185 0.02 0.226 0.004 953.48 7,567.87 326.03 BULQIZË TREBISHT 993 0.179 0.04 0.233 0.010 231.11 7,752.53 721.70 BULQIZË OSTREN 3,034 0.202 0.04 0.242 0.008 733.17 7,636.99 637.82 BULQIZË SHUPENZË 5,499 0.195 0.03 0.229 0.006 1,257.73 7,538.49 552.88 BULQIZË ZERQAN 4,111 0.239 0.05 0.232 0.011 952.50 7,154.24 689.88 BULQIZË MARTANESH 1,827 0.144 0.08 0.247 0.018 451.03 8,478.82 982.58 DELVINË DELVINË 5,754 0.130 0.02 0.255 0.005 1,467.59 9,060.40 454.50 DELVINË FINIQ 1,333 0.120 0.05 0.248 0.010 330.78 9,055.23 814.25 DELVINË MESOPOTAM 2,786 0.119 0.03 0.247 0.007 688.13 9,059.55 485.26 DELVINË VERGO 1,844 0.152 0.05 0.259 0.011 478.33 8,772.81 693.48 DEVOLL BILISHT 6,250 0.085 0.06 0.265 0.014 1,655.37 10,510.27 682.51 DEVOLL QENDËR BILISHT 5,440 0.107 0.07 0.255 0.017 1,387.33 9,505.65 723.17 DEVOLL MIRAS 6,577 0.096 0.06 0.248 0.015 1,632.94 9,549.12 631.64 DEVOLL HOÇISHT 4,465 0.107 0.05 0.253 0.013 1,131.13 9,436.70 462.01 DEVOLL PROGËR 3,988 0.094 0.03 0.251 0.007 1,001.63 9,717.08 384.48 DIBËR ARRAS 3,055 0.233 0.04 0.218 0.010 666.40 7,000.51 472.05 DIBËR FUSHË-ÇIDHËN 2,909 0.190 0.04 0.218 0.010 633.94 7,381.44 480.09 DIBËR KALAEDODËS 2,252 0.250 0.06 0.204 0.015 458.49 6,721.67 577.87 DIBËR KASTRIOT 6,201 0.208 0.02 0.226 0.005 1,401.04 7,338.83 400.23 DIBËR LURË 1,096 0.197 0.05 0.242 0.011 265.45 7,671.74 459.02 DIBËR MAQELLARË 10,666 0.240 0.05 0.225 0.011 2,400.35 7,051.03 516.77 DIBËR MELAN 3,649 0.216 0.05 0.212 0.012 775.32 7,088.48 688.29 DIBËR FUSHË-MUHUR 2,780 0.192 0.07 0.228 0.018 635.05 7,554.82 623.59 DIBËR PESHKOPI 13,251 0.207 0.02 0.234 0.004 3,102.99 7,524.95 486.32 DIBËR QENDËR TOMIN 7,591 0.195 0.05 0.222 0.012 1,684.63 7,406.90 609.35 DIBËR SELISHTË 1,605 0.264 0.02 0.217 0.005 348.49 6,765.42 320.28 DIBËR SLLOVË 2,405 0.294 0.02 0.224 0.005 537.88 6,640.84 308.91 DIBËR ZALL DARDHË 1,045 0.289 0.02 0.252 0.006 263.00 6,954.80 370.42 DIBËR ZALLREÇ 681 0.284 0.03 0.218 0.007 148.28 6,659.97 367.84 DIBËR LUZNI 2,433 0.259 0.02 0.216 0.004 526.53 6,781.79 384.30 DURRËS DURRËS 112,957 0.141 0.04 0.279 0.009 31,532.36 9,350.70 383.86 DURRËS GJEPALAJ 3,449 0.197 0.03 0.236 0.007 815.11 7,658.84 404.03 DURRËS ISHËM 5,001 0.170 0.01 0.243 0.003 1,215.02 8,073.33 186.36 DURRËS KATUND I RI 10,161 0.202 0.05 0.238 0.012 2,422.19 7,606.73 743.86 Portraits of Poverty and Inequality in Albania | Annexes 33 Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption DURRËS MAMINAS 4,458 0.134 0.05 0.250 0.012 1,115.93 8,777.43 591.58 DURRËS MANËZ 6,652 0.209 0.02 0.251 0.005 1,670.59 7,787.06 382.27 DURRËS RRASHBULL 23,952 0.167 0.04 0.247 0.010 5,911.82 8,167.22 509.30 DURRËS SHIJAK 7,573 0.125 0.02 0.266 0.005 2,011.28 9,381.52 344.01 DURRËS SUKTH 15,966 0.211 0.06 0.238 0.015 3,803.94 7,514.20 682.01 DURRËS XHAFZOTAJ 12,381 0.160 0.03 0.249 0.008 3,079.83 8,325.25 517.96 ELBASAN BELSH 8,781 0.092 0.03 0.243 0.006 2,132.61 9,481.24 442.46 ELBASAN BRADASHESH 10,700 0.173 0.03 0.239 0.007 2,560.68 8,024.96 432.87 ELBASAN CËRRIK 6,695 0.147 0.04 0.257 0.009 1,721.89 8,805.69 394.97 ELBASAN ELBASAN 78,467 0.127 0.07 0.263 0.018 20,638.38 9,323.16 597.56 ELBASAN FIERZË 2,065 0.129 0.04 0.246 0.009 509.01 8,956.70 408.28 ELBASAN FUNARË 2,122 0.175 0.03 0.233 0.008 493.47 7,900.93 483.71 ELBASAN GJERGJAN 5,126 0.116 0.05 0.240 0.012 1,232.14 8,914.58 601.57 ELBASAN GJINAR 3,233 0.153 0.03 0.244 0.007 787.97 8,384.91 446.98 ELBASAN GOSTIMË 8,116 0.119 0.02 0.248 0.005 2,009.75 9,043.50 363.93 ELBASAN GRACEN 2,192 0.173 0.03 0.241 0.008 529.20 8,152.79 380.41 ELBASAN GREKAN 3,138 0.130 0.04 0.244 0.010 764.93 8,734.89 440.66 ELBASAN KAJAN 3,925 0.123 0.06 0.239 0.015 937.52 8,766.91 755.72 ELBASAN KLOS 3,262 0.133 0.03 0.236 0.007 769.54 8,526.69 381.20 ELBASAN LABINOT-FUSHË 7,058 0.180 0.05 0.234 0.012 1,649.48 7,838.76 684.11 ELBASAN LABINOT-MAL 5,291 0.232 0.03 0.222 0.006 1,174.89 7,133.00 390.75 ELBASAN MOLLAS 5,530 0.179 0.03 0.237 0.009 1,309.16 7,913.17 474.26 ELBASAN PAPËR 6,348 0.133 0.03 0.235 0.007 1,490.23 8,523.49 489.17 ELBASAN RRASË 1,594 0.146 0.03 0.230 0.008 366.89 8,234.83 445.28 ELBASAN SHALËS 3,842 0.132 0.03 0.247 0.007 950.63 8,857.88 1,181.04 ELBASAN SHIRGJAN 7,307 0.116 0.03 0.247 0.005 1,808.13 9,101.66 683.18 ELBASAN SHUSHICË 8,722 0.156 0.03 0.234 0.006 2,043.71 8,142.32 640.93 ELBASAN TREGAN 3,036 0.166 0.02 0.232 0.005 703.00 7,967.98 593.76 ELBASAN ZAVALIN 1,622 0.162 0.01 0.246 0.003 399.68 8,394.10 386.01 FIER CAKRAN 11,722 0.194 0.03 0.237 0.007 2,774.67 7,659.54 944.36 FIER DERMENAS 7,788 0.188 0.03 0.237 0.006 1,845.73 7,748.79 979.97 FIER FIER 55,752 0.110 0.02 0.271 0.004 15,093.48 9,826.79 736.23 FIER FRAKULL 6,820 0.161 0.03 0.234 0.007 1,593.22 8,011.48 1,378.06 FIER RUZHDIE 2,328 0.179 0.04 0.220 0.008 512.60 7,597.36 739.66 FIER KUMAN 5,611 0.194 0.03 0.235 0.005 1,319.16 7,660.11 1,437.85 FIER KURJAN 3,618 0.204 0.04 0.228 0.007 825.37 7,436.27 956.02 FIER LEVAN 8,159 0.162 0.02 0.246 0.005 2,011.01 8,269.19 1,207.65 FIER LIBOFSHË 6,149 0.151 0.04 0.245 0.008 1,505.68 8,418.25 776.19 FIER MBROSTAR 7,460 0.193 0.03 0.241 0.007 1,794.37 7,753.67 757.50 FIER PATOS 15,397 0.144 0.06 0.246 0.013 3,792.67 8,556.48 840.06 FIER PORTËZ 8,259 0.164 0.03 0.235 0.007 1,940.90 7,992.24 701.85 FIER QENDËR 4,207 0.168 0.02 0.241 0.004 1,014.89 8,103.62 510.13 FIER ROSKOVEC 4,975 0.150 0.03 0.245 0.006 1,219.48 8,443.80 623.15 FIER STRUM 7,538 0.208 0.03 0.222 0.006 1,676.02 7,286.74 704.58 FIER TOPOJË 4,246 0.157 0.03 0.247 0.007 1,047.42 8,385.20 571.36 FIER ZHARRËS 5,247 0.166 0.03 0.245 0.007 1,283.03 8,167.52 615.37 GRAMSH GRAMSH 8,440 0.123 0.05 0.240 0.010 2,028.71 8,750.16 683.07 GRAMSH KODOVJAT 2,355 0.173 0.04 0.226 0.008 531.12 7,691.24 706.28 GRAMSH KUSHOVË 659 0.118 0.04 0.224 0.010 147.56 8,424.68 439.94 GRAMSH LENIE 779 0.157 0.06 0.216 0.013 168.54 7,748.44 756.97 GRAMSH PISHAJ 4,906 0.200 0.06 0.233 0.014 1,145.49 7,552.86 827.59 GRAMSH POROÇAN 1,269 0.268 0.04 0.229 0.008 290.28 6,868.46 702.41 34 Portraits of Poverty and Inequality in Albania | Annexes Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption GRAMSH SKËNDERBEGAS 1,239 0.175 0.03 0.249 0.008 308.80 8,101.17 479.74 GRAMSH KUKUR 2,560 0.123 0.03 0.239 0.007 612.85 8,564.93 541.14 GRAMSH SULT 631 0.202 0.05 0.218 0.012 137.38 7,338.24 691.66 GRAMSH TUNJË 1,393 0.193 0.02 0.230 0.005 320.52 7,593.94 597.05 GJIROKASTËR ANTIGONE 998 0.053 0.02 0.260 0.005 259.23 11,649.14 566.55 GJIROKASTËR CEPO 1,727 0.070 0.02 0.255 0.005 440.22 10,567.58 536.86 GJIROKASTËR DROPULL I POSHTËM 2,095 0.096 0.02 0.254 0.004 532.83 9,932.79 339.50 GJIROKASTËR DROPULL I SIPËRM 969 0.061 0.02 0.261 0.005 252.82 11,536.34 366.34 GJIROKASTËR GJIROKASTËR 19,708 0.069 0.02 0.272 0.005 5,352.86 11,295.29 405.05 GJIROKASTËR LAZARAT 2,801 0.070 0.05 0.268 0.011 750.90 11,195.05 1,096.89 GJIROKASTËR LIBOHOVË 1,992 0.064 0.06 0.249 0.013 496.55 10,736.00 800.93 GJIROKASTËR LUNXHËRI 1,941 0.052 0.03 0.252 0.005 488.20 11,227.19 734.71 GJIROKASTËR ODRIE 433 0.039 0.04 0.242 0.009 104.66 11,838.65 825.41 GJIROKASTËR PICAR 937 0.088 0.05 0.252 0.011 236.52 10,040.16 1,202.51 GJIROKASTËR POGON 432 0.031 0.03 0.228 0.006 98.54 11,935.76 809.77 GJIROKASTËR QENDËR LIBOHOVË 1,264 0.085 0.06 0.257 0.013 324.67 10,401.89 938.23 GJIROKASTËR ZAGORI 411 0.026 0.03 0.240 0.005 98.67 13,215.98 504.54 HAS FAJZA 3,476 0.258 0.02 0.229 0.005 794.51 6,954.45 440.03 HAS GOLAJ 6,187 0.231 0.04 0.237 0.009 1,469.39 7,311.64 593.26 HAS GJINAJ 1,106 0.257 0.01 0.213 0.003 236.09 6,806.49 276.92 HAS KRUMË 6,006 0.167 0.06 0.235 0.014 1,412.20 7,990.25 772.38 KAVAJË GOLEM 6,833 0.171 0.02 0.253 0.004 1,729.25 8,247.80 352.45 KAVAJË GOSË 4,120 0.164 0.02 0.246 0.003 1,011.95 8,224.78 555.14 KAVAJË HELMËS 3,139 0.198 0.04 0.234 0.009 733.36 7,596.85 539.60 KAVAJË KAVAJË 20,144 0.151 0.05 0.264 0.011 5,321.64 8,897.86 728.36 KAVAJË KRYEVIDH 4,662 0.174 0.03 0.242 0.006 1,129.33 8,050.43 481.88 KAVAJË LEKAJ 5,116 0.160 0.02 0.240 0.005 1,225.28 8,165.05 295.39 KAVAJË LUZIVOGEL 4,735 0.136 0.02 0.246 0.004 1,163.58 8,656.10 309.03 KAVAJË RROGOZHINË 6,650 0.209 0.02 0.251 0.004 1,666.04 7,769.04 392.02 KAVAJË SINABALLAJ 1,191 0.220 0.03 0.215 0.008 256.45 7,152.35 578.22 KAVAJË SYNEJ 4,995 0.165 0.03 0.252 0.007 1,258.43 8,385.83 444.64 KOLONJË BARMASH 480 0.077 0.07 0.235 0.017 112.86 9,994.94 1,057.55 KOLONJË ÇLIRIM 355 0.113 0.03 0.226 0.006 80.24 8,769.87 367.16 KOLONJË ERSEKË 3,746 0.071 0.02 0.246 0.005 921.10 10,266.23 249.81 KOLONJË LESKOVIK 1,525 0.089 0.04 0.238 0.011 362.81 9,667.71 347.69 KOLONJË LESKOVIK 416 0.087 0.04 0.231 0.010 96.12 9,786.44 554.13 KOLONJË MOLLAS 1,520 0.073 0.02 0.246 0.004 373.99 10,195.94 282.86 KOLONJË NOVOSELË 355 0.087 0.03 0.233 0.008 82.80 9,490.57 375.55 KOLONJË QENDËR 2,673 0.083 0.03 0.255 0.008 682.88 10,151.93 390.91 KORÇË DRENOVË 5,199 0.100 0.05 0.259 0.011 1,344.94 9,752.06 529.57 KORÇË GORË 1,565 0.119 0.04 0.247 0.010 386.03 9,075.09 557.84 KORÇË KORÇË 50,847 0.096 0.06 0.270 0.014 13,727.29 10,393.99 444.98 KORÇË LEKAS 386 0.143 0.06 0.245 0.015 94.60 8,626.25 677.32 KORÇË LIBONIK 8,922 0.116 0.05 0.247 0.012 2,204.61 9,101.16 503.79 KORÇË LIQENAS 3,154 0.051 0.05 0.249 0.012 786.32 11,140.89 575.27 KORÇË MALIQ 4,290 0.122 0.02 0.242 0.005 1,040.17 8,902.12 418.94 KORÇË MOGLICË 951 0.123 0.04 0.237 0.009 225.15 8,845.02 508.40 KORÇË MOLLAJ 3,438 0.113 0.08 0.249 0.018 854.55 9,226.25 736.11 KORÇË PIRG 7,652 0.138 0.03 0.241 0.007 1,845.00 8,551.97 331.51 KORÇË POJAN 10,864 0.115 0.03 0.253 0.006 2,747.14 9,300.05 382.16 KORÇË QENDËR 9,022 0.103 0.05 0.250 0.011 2,257.93 9,451.12 486.20 KORÇË VITHKUQ 1,510 0.110 0.03 0.248 0.007 374.83 9,312.07 730.47 Portraits of Poverty and Inequality in Albania | Annexes 35 Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption KORÇË VOSKOP 3,832 0.130 0.03 0.251 0.007 963.29 8,941.19 585.97 KORÇË VOSKOPOJË 1,058 0.130 0.04 0.270 0.009 285.87 9,560.73 456.59 KORÇË VRESHTAS 7,641 0.126 0.09 0.237 0.024 1,807.66 8,629.27 780.31 KRUJË BUBQ 5,874 0.090 0.10 0.252 0.026 1,480.02 9,788.12 759.77 KRUJË CUDHI 1,812 0.160 0.06 0.238 0.015 432.08 8,218.37 644.12 KRUJË FUSHË-KRUJË 18,102 0.146 0.05 0.250 0.013 4,518.06 8,618.52 596.91 KRUJË KODËR-THUMANË 12,334 0.141 0.10 0.253 0.024 3,119.99 8,728.76 953.74 KRUJË KRUJË 11,654 0.134 0.04 0.254 0.009 2,965.58 8,974.38 621.36 KRUJË NIKËL 9,518 0.139 0.08 0.249 0.020 2,371.77 8,732.18 653.47 KUÇOVË KOZARE 5,622 0.125 0.06 0.241 0.015 1,353.48 8,768.89 570.63 KUÇOVË KUÇOVË 12,650 0.122 0.07 0.253 0.018 3,194.31 9,211.03 665.96 KUÇOVË PERONDI 9,011 0.108 0.05 0.245 0.013 2,204.54 9,175.43 656.67 KUKËS ARRËN 462 0.227 0.04 0.215 0.009 99.19 7,118.79 593.17 KUKËS BICAJ 5,631 0.256 0.04 0.219 0.009 1,233.41 6,856.62 543.23 KUKËS BUSHTRICË 1,486 0.255 0.02 0.216 0.004 321.59 6,884.21 270.90 KUKËS KOLSH 1,250 0.328 0.02 0.220 0.004 274.56 6,422.92 405.26 KUKËS KUKËS 16,697 0.170 0.03 0.230 0.007 3,846.84 7,824.08 373.47 KUKËS MALZI 3,072 0.240 0.04 0.220 0.010 675.67 6,980.26 454.51 KUKËS ZAPOD 2,217 0.293 0.03 0.218 0.007 483.16 6,583.69 386.25 KUKËS SHISHTAVEC 3,835 0.265 0.03 0.235 0.008 902.38 6,974.33 447.37 KUKËS SHTIQËN 3,438 0.175 0.08 0.225 0.018 774.65 7,729.87 1,031.27 KUKËS SURROJ 1,099 0.356 0.02 0.209 0.005 229.40 6,123.54 219.94 KUKËS TËRTHORE 2,959 0.164 0.07 0.217 0.016 642.77 7,683.22 1,513.54 KUKËS TOPOJAN 1,753 0.337 0.01 0.208 0.003 365.20 6,200.18 222.57 KUKËS UJËMISHT 1,797 0.365 0.08 0.202 0.019 362.72 5,982.50 1,090.47 KUKËS GRYKËÇAJË 1,440 0.300 0.10 0.199 0.021 286.80 6,337.24 1,659.11 KUKËS KALIS 827 0.385 0.02 0.197 0.005 162.84 5,838.82 424.52 KURBIN FUSHË-KUQE 5,450 0.200 0.06 0.245 0.015 1,336.63 7,759.97 655.51 KURBIN LAÇ 17,067 0.188 0.07 0.249 0.017 4,247.06 7,939.04 918.19 KURBIN MAMURRAS 15,280 0.198 0.03 0.236 0.008 3,601.49 7,591.08 465.40 KURBIN MILOT 8,452 0.191 0.03 0.239 0.007 2,017.21 7,725.84 417.56 LEZHË BALLDRENIRI 6,138 0.201 0.03 0.242 0.008 1,485.98 7,694.39 495.51 LEZHË BLINISHT 3,348 0.163 0.02 0.244 0.005 815.51 8,171.05 317.24 LEZHË DAJÇ 3,826 0.154 0.01 0.250 0.003 957.72 8,474.26 248.02 LEZHË KALLMET 4,101 0.157 0.02 0.235 0.006 962.20 8,092.07 447.49 LEZHË KOLÇ 4,225 0.168 0.03 0.248 0.008 1,049.89 8,199.39 280.69 LEZHË LEZHË 15,496 0.130 0.03 0.287 0.006 4,451.21 9,811.57 564.48 LEZHË SHËNGJIN 8,087 0.138 0.03 0.268 0.008 2,169.05 9,128.28 439.02 LEZHË SHËNKOLL 12,474 0.192 0.03 0.241 0.007 3,000.97 7,734.43 604.78 LEZHË UNGREJ 1,587 0.202 0.02 0.221 0.004 350.97 7,368.76 388.84 LEZHË ZEJMEN 5,656 0.174 0.05 0.241 0.010 1,362.84 7,984.80 607.43 LIBRAZHD HOTOLISHT 5,700 0.220 0.03 0.215 0.008 1,227.35 7,082.10 397.87 LIBRAZHD LIBRAZHD 6,937 0.087 0.05 0.233 0.012 1,619.48 9,268.62 744.31 LIBRAZHD LUNIK 2,621 0.245 0.04 0.244 0.009 639.22 7,252.61 373.11 LIBRAZHD ORENJË 3,883 0.304 0.03 0.225 0.006 872.11 6,562.22 333.54 LIBRAZHD PËRRENJAS 5,846 0.162 0.06 0.240 0.014 1,404.01 8,125.88 422.15 LIBRAZHD POLIS 3,385 0.205 0.04 0.235 0.009 796.14 7,525.33 337.34 LIBRAZHD QENDËR 8,557 0.195 0.07 0.232 0.015 1,982.73 7,512.78 677.06 LIBRAZHD QUKËS 8,211 0.215 0.04 0.228 0.010 1,868.49 7,296.68 427.84 LIBRAZHD STËBLEVË 809 0.122 0.04 0.237 0.010 192.10 8,665.03 364.54 LIBRAZHD STRAVAJ 2,427 0.206 0.05 0.239 0.012 580.05 7,563.11 386.26 LIBRAZHD RRAJCË 8,421 0.224 0.05 0.229 0.010 1,925.70 7,227.01 539.32 36 Portraits of Poverty and Inequality in Albania | Annexes Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption LUSHNJE ALLKAJ 4,319 0.169 0.04 0.236 0.011 1,019.59 7,978.10 321.30 LUSHNJE BALLAGAT 2,461 0.192 0.05 0.223 0.010 547.58 7,515.81 421.36 LUSHNJE BUBULLIMË 5,548 0.144 0.06 0.238 0.014 1,321.26 8,343.01 373.93 LUSHNJE DIVJAKË 8,453 0.169 0.04 0.249 0.009 2,101.49 8,233.48 298.35 LUSHNJE DUSHK 7,872 0.199 0.06 0.229 0.014 1,801.46 7,499.75 552.99 LUSHNJE FIER-SHEGAN 7,023 0.172 0.03 0.229 0.008 1,608.99 7,792.52 191.84 LUSHNJE GOLEM 5,240 0.197 0.04 0.233 0.010 1,220.32 7,608.24 332.18 LUSHNJE GRABIAN 3,638 0.201 0.04 0.222 0.010 807.25 7,417.91 379.35 LUSHNJE GRADISHTË 7,521 0.149 0.03 0.242 0.008 1,818.90 8,336.68 300.83 LUSHNJE HYSGJOKAJ 2,603 0.191 0.03 0.220 0.008 573.21 7,474.75 279.41 LUSHNJE KARBUNARË 4,193 0.203 0.06 0.231 0.015 969.68 7,507.21 407.59 LUSHNJE KOLONJË 5,728 0.136 0.06 0.256 0.017 1,463.67 8,846.57 379.46 LUSHNJE KRUTJE 7,564 0.148 0.07 0.234 0.019 1,767.97 8,182.93 479.83 LUSHNJE LUSHNJE 31,093 0.125 0.08 0.265 0.022 8,240.47 9,343.56 539.67 LUSHNJE RREMAS 4,449 0.189 0.06 0.229 0.015 1,018.63 7,613.55 400.62 LUSHNJE TËRBUF 10,201 0.216 0.02 0.233 0.005 2,377.36 7,390.08 344.04 MALËSI E MADHE GRUEMIRË 8,887 0.163 0.03 0.255 0.007 2,263.65 8,444.37 332.35 MALËSI E MADHE KASTRAT 6,877 0.148 0.06 0.263 0.011 1,810.97 8,877.81 683.45 MALËSI E MADHE KELMEND 3,056 0.170 0.06 0.250 0.012 764.37 8,297.15 547.41 MALËSI E MADHE KOPLIK 3,734 0.093 0.03 0.264 0.008 987.22 10,134.42 347.66 MALËSI E MADHE QENDËR 4,740 0.123 0.06 0.254 0.015 1,205.64 9,154.78 411.08 MALËSI E MADHE SHKREL 3,520 0.181 0.04 0.258 0.010 908.30 8,283.57 507.53 MALLAKASTËR ARANITAS 2,709 0.196 0.03 0.223 0.006 605.17 7,458.58 459.33 MALLAKASTËR BALLSH 7,657 0.114 0.06 0.254 0.014 1,946.21 9,251.07 509.53 MALLAKASTËR FRATAR 3,435 0.201 0.05 0.232 0.010 797.68 7,519.99 407.66 MALLAKASTËR GRESHICË 1,152 0.173 0.05 0.230 0.011 265.51 7,841.83 365.17 MALLAKASTËR HEKAL 2,623 0.172 0.02 0.237 0.004 620.78 7,971.02 414.87 MALLAKASTËR KUTË 1,977 0.165 0.04 0.236 0.011 466.64 8,008.70 348.26 MALLAKASTËR NGRAÇAN 588 0.188 0.05 0.208 0.013 122.09 7,468.07 314.08 MALLAKASTËR QENDËR 6,253 0.178 0.03 0.233 0.007 1,455.17 7,784.03 349.95 MALLAKASTËR SELITË 877 0.221 0.04 0.233 0.010 204.24 7,426.92 399.23 MAT BAZ 2,228 0.187 0.03 0.235 0.007 523.88 7,838.82 283.15 MAT DERJAN 1,098 0.180 0.03 0.231 0.006 253.74 7,791.28 239.81 MAT GURRË 3,369 0.232 0.05 0.232 0.011 782.08 7,253.27 746.34 MAT KLOS 7,873 0.159 0.04 0.242 0.010 1,908.92 8,246.97 414.34 MAT KOMSI 4,283 0.177 0.03 0.244 0.007 1,044.01 8,070.33 237.43 MAT LIS 3,824 0.168 0.05 0.248 0.012 947.11 8,238.40 334.48 MAT MACUKULL 1,565 0.194 0.04 0.231 0.010 362.15 7,644.82 292.45 MAT BURREL 10,664 0.112 0.08 0.243 0.021 2,596.25 9,101.31 540.79 MAT RUKAJ 2,511 0.186 0.04 0.236 0.009 592.08 7,794.22 447.16 MAT SUÇ 2,716 0.173 0.10 0.241 0.022 655.04 8,066.61 710.86 MAT ULËZ 1,229 0.152 0.05 0.250 0.011 307.75 8,602.79 309.19 MAT XIBËR 2,660 0.204 0.06 0.238 0.016 633.68 7,659.66 480.41 MIRDITË FAN 2,974 0.238 0.10 0.236 0.031 703.30 7,268.04 580.12 MIRDITË KAÇINAR 1,016 0.180 0.03 0.246 0.006 250.34 8,109.76 282.02 MIRDITË KTHJELLË 2,206 0.181 0.05 0.243 0.012 535.66 7,962.58 346.28 MIRDITË OROSH 1,899 0.155 0.07 0.247 0.017 469.68 8,435.14 437.85 MIRDITË RRËSHEN 8,796 0.112 0.04 0.248 0.011 2,185.17 9,179.91 285.90 MIRDITË RUBIK 4,452 0.158 0.05 0.249 0.011 1,107.54 8,432.30 504.61 MIRDITË SELITË 745 0.189 0.10 0.242 0.028 180.61 7,913.32 516.02 PEQIN GJOCAJ 5,207 0.149 0.04 0.235 0.009 1,223.27 8,268.61 411.18 PEQIN KARINË 1,350 0.144 0.08 0.234 0.022 315.29 8,370.13 401.20 Portraits of Poverty and Inequality in Albania | Annexes 37 Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption PEQIN PAJOVË 6,626 0.137 0.08 0.244 0.025 1,613.89 8,605.66 428.39 PEQIN PEQIN 5,616 0.148 0.08 0.264 0.022 1,484.53 8,993.70 469.67 PEQIN PËRPARIM 3,423 0.125 0.10 0.240 0.030 820.31 8,794.67 484.75 PEQIN SHEZË 3,177 0.134 0.02 0.245 0.005 778.94 8,730.98 496.63 PËRMET BALLABAN 1,047 0.093 0.04 0.245 0.010 256.04 9,598.28 404.71 PËRMET ÇARÇOVË 910 0.067 0.04 0.247 0.008 224.66 10,640.84 568.53 PËRMET FRASHËR 387 0.116 0.05 0.243 0.011 93.88 9,083.72 476.19 PËRMET KËLCYRË 2,651 0.091 0.08 0.243 0.023 644.20 9,700.74 475.89 PËRMET PËRMET 5,945 0.070 0.07 0.248 0.016 1,472.49 10,359.98 496.43 PËRMET QENDËR 1,742 0.081 0.05 0.245 0.012 427.26 9,967.17 460.47 PËRMET SUKË 1,256 0.075 0.05 0.235 0.012 295.29 9,819.81 318.16 PËRMET DISHNICË 1,159 0.114 0.01 0.246 0.003 285.10 9,220.09 180.12 PËRMET PETRAN 1,622 0.092 0.05 0.247 0.012 400.99 9,750.14 444.24 POGRADEC BUÇIMAS 15,687 0.156 0.03 0.247 0.008 3,873.35 8,374.72 370.37 POGRADEC ÇËRRAVË 7,009 0.211 0.03 0.236 0.008 1,653.63 7,527.11 303.32 POGRADEC DARDHAS 2,182 0.156 0.03 0.243 0.008 529.30 8,382.61 461.39 POGRADEC POGRADEC 20,851 0.108 0.04 0.260 0.011 5,428.39 9,676.43 368.22 POGRADEC PROPTISHT 4,782 0.164 0.02 0.236 0.005 1,126.69 8,067.55 244.02 POGRADEC TREBINJË 2,481 0.147 0.03 0.232 0.006 574.39 8,225.60 436.63 POGRADEC HUDENISHT 5,990 0.170 0.03 0.240 0.008 1,438.58 8,118.53 273.41 POGRADEC VELÇAN 2,546 0.143 0.02 0.236 0.006 602.12 8,388.82 307.24 PUKË BLERIM 913 0.237 0.03 0.240 0.006 218.95 7,431.05 236.11 PUKË FIERZË 1,302 0.221 0.03 0.233 0.008 303.13 7,424.98 295.80 PUKË FUSHË-ARRËZ 2,513 0.187 0.01 0.245 0.002 616.10 8,022.87 200.14 PUKË GJEGJAN 2,846 0.174 0.03 0.242 0.007 687.34 8,040.62 318.38 PUKË IBALLË 1,129 0.184 0.05 0.239 0.011 269.52 7,974.87 484.03 PUKË PUKË 3,603 0.126 0.04 0.249 0.009 895.75 8,976.58 320.28 PUKË QELËZ 1,761 0.221 0.04 0.226 0.010 398.41 7,286.73 342.08 PUKË QERRET 1,486 0.195 0.03 0.243 0.006 361.55 7,853.75 303.88 PUKË QAFË MALI 1,534 0.212 0.03 0.247 0.006 379.59 7,690.50 327.01 PUKË RRAPË 1,357 0.154 0.03 0.244 0.009 331.20 8,386.03 314.48 SARANDË DHIVËR 1,396 0.068 0.02 0.236 0.005 330.00 10,502.49 290.50 SARANDË KONISPOL 2,123 0.152 0.03 0.261 0.007 554.72 8,818.00 297.01 SARANDË LIVADHJA 1,165 0.059 0.04 0.239 0.009 278.21 10,843.44 451.19 SARANDË LUKOVË 2,916 0.128 0.03 0.265 0.008 773.17 9,416.50 456.87 SARANDË SARANDË 17,161 0.107 0.03 0.278 0.008 4,766.52 10,133.88 271.47 SARANDË XARRË 4,254 0.162 0.03 0.227 0.008 964.22 7,880.78 420.08 SARANDË ALIKO 3,849 0.109 0.03 0.264 0.007 1,017.41 9,766.03 329.80 SARANDË MARKAT 1,859 0.136 0.04 0.241 0.008 447.31 8,663.92 364.50 SARANDË KSAMIL 3,003 0.139 0.06 0.247 0.014 741.52 8,641.61 568.64 SKRAPAR BOGOVË 1,088 0.110 0.03 0.241 0.007 261.70 9,189.55 374.53 SKRAPAR ÇEPAN 740 0.078 0.03 0.240 0.007 177.61 9,823.77 347.59 SKRAPAR ÇOROVODË 4,051 0.099 0.04 0.242 0.009 982.30 9,447.71 336.13 SKRAPAR GJERBËS 813 0.116 0.03 0.234 0.007 190.19 8,833.49 327.58 SKRAPAR LESHNJË 496 0.131 0.05 0.232 0.011 115.13 8,636.08 485.26 SKRAPAR POLIÇAN 4,309 0.100 0.06 0.234 0.015 1,008.38 9,149.34 609.78 SKRAPAR POTOM 897 0.096 0.03 0.222 0.006 199.02 8,931.75 311.52 SKRAPAR QENDËR 2,545 0.105 0.06 0.239 0.014 607.06 9,158.69 529.30 SKRAPAR VENDRESHË 984 0.131 0.05 0.234 0.011 230.26 8,625.20 424.16 SKRAPAR ZHEPË 779 0.169 0.03 0.240 0.006 187.12 8,111.39 369.25 SHKODËR ANAEMALIT 3,858 0.106 0.03 0.260 0.006 1,002.52 9,741.27 301.48 SHKODËR BËRDICË 5,775 0.148 0.01 0.252 0.003 1,457.39 8,690.74 225.91 38 Portraits of Poverty and Inequality in Albania | Annexes Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption SHKODËR BUSHAT 14,102 0.121 0.04 0.265 0.009 3,739.39 9,436.11 333.35 SHKODËR DAJÇ 3,880 0.091 0.03 0.266 0.008 1,030.91 10,269.55 267.83 SHKODËR GURIIZI 8,067 0.142 0.05 0.251 0.012 2,028.16 8,754.40 447.60 SHKODËR HAJMEL 4,415 0.189 0.03 0.255 0.006 1,123.69 8,128.08 436.75 SHKODËR VAU I DEJËS 8,114 0.146 0.05 0.254 0.010 2,057.86 8,715.98 383.46 SHKODËR POSTRIBË 7,059 0.159 0.07 0.251 0.015 1,772.67 8,426.24 645.61 SHKODËR PULT 1,529 0.164 0.04 0.248 0.009 378.54 8,465.82 426.07 SHKODËR RRETHINAT 21,133 0.170 0.04 0.258 0.010 5,454.11 8,384.77 454.09 SHKODËR SHALË 1,795 0.087 0.10 0.259 0.026 465.40 10,316.30 954.63 SHKODËR SHKODËR 75,649 0.122 0.03 0.269 0.007 20,354.73 9,621.64 304.15 SHKODËR SHLLAK 680 0.135 0.09 0.247 0.021 168.25 8,878.07 694.89 SHKODËR SHOSH 309 0.102 0.05 0.250 0.012 77.35 9,684.60 462.15 SHKODËR VELIPOJË 5,015 0.115 0.03 0.263 0.006 1,318.07 9,473.51 265.42 SHKODËR VIG-MNELË 1,509 0.190 0.03 0.246 0.007 370.62 7,996.32 276.45 SHKODËR TEMAL 1,559 0.142 0.04 0.244 0.008 380.42 8,653.39 351.92 TEPELENË BUZ 737 0.106 0.05 0.243 0.012 178.73 9,299.69 443.18 TEPELENË FSHAT MEMALIAJ 1,606 0.093 0.04 0.241 0.009 387.66 9,518.99 414.77 TEPELENË KRAHËS 2,444 0.143 0.03 0.250 0.007 610.74 8,751.06 413.46 TEPELENË KURVELESH 705 0.129 0.04 0.255 0.008 179.88 9,148.02 431.93 TEPELENË LOPËS 723 0.133 0.04 0.236 0.009 170.84 8,760.87 434.45 TEPELENË LUFTINJË 1,734 0.106 0.02 0.245 0.005 425.64 9,401.37 348.44 TEPELENË MEMALIAJ 2,647 0.120 0.03 0.243 0.006 643.77 9,019.06 424.98 TEPELENË QENDËR 3,119 0.116 0.03 0.250 0.007 780.11 9,224.39 269.60 TEPELENË QESARAT 1,379 0.136 0.07 0.253 0.017 349.12 8,967.59 585.41 TEPELENË TEPELENË 4,342 0.081 0.03 0.255 0.008 1,105.77 10,278.72 355.50 TIRANË BALDUSHK 4,576 0.132 0.03 0.234 0.007 1,071.69 8,511.35 340.08 TIRANË BËRXULLË 11,030 0.145 0.03 0.248 0.008 2,740.77 8,631.82 428.55 TIRANË BËRZHITË 4,973 0.139 0.05 0.243 0.011 1,208.45 8,610.84 410.64 TIRANË DAJT 20,139 0.126 0.02 0.280 0.005 5,638.56 9,732.73 267.31 TIRANË KAMËZ 66,832 0.252 0.03 0.253 0.008 16,903.52 7,339.52 313.26 TIRANË KASHAR 43,058 0.112 0.03 0.269 0.007 11,584.31 9,791.84 363.78 TIRANË NDROQ 5,035 0.112 0.03 0.251 0.007 1,265.77 9,264.34 412.60 TIRANË PASKUQAN 37,305 0.209 0.04 0.243 0.010 9,058.32 7,645.58 370.72 TIRANË PETRELË 5,542 0.102 0.08 0.251 0.019 1,393.10 9,492.32 654.57 TIRANË PEZË 6,272 0.138 0.03 0.244 0.008 1,532.12 8,650.64 366.86 TIRANË PREZË 4,727 0.103 0.03 0.265 0.006 1,251.91 9,966.06 413.62 TIRANË FARKË 22,614 0.098 0.04 0.300 0.010 6,776.34 11,001.49 566.13 TIRANË SHËNGJERGJ 2,186 0.146 0.03 0.235 0.007 514.32 8,323.45 539.41 TIRANË TIRANË 410,813 0.092 0.04 0.277 0.009 113,752.32 10,647.36 540.19 TIRANË VAQARR 8,922 0.149 0.03 0.249 0.005 2,224.76 8,557.52 658.85 TIRANË VORË 10,873 0.186 0.05 0.256 0.013 2,785.26 8,157.85 753.00 TIRANË ZALLBASTAR 3,380 0.147 0.03 0.232 0.006 784.42 8,271.41 750.09 TIRANË ZALLHERR 9,389 0.172 0.03 0.235 0.007 2,209.42 7,969.64 521.21 TIRANË KËRRABË 3,264 0.249 0.02 0.242 0.004 790.64 7,236.50 342.26 TROPOJË BAJRAM CURRI 5,337 0.100 0.04 0.249 0.010 1,327.76 9,405.96 503.45 TROPOJË BUJAN 2,550 0.208 0.03 0.244 0.006 622.51 7,639.10 437.75 TROPOJË BYTYÇ 1,563 0.116 0.05 0.243 0.011 379.62 8,757.64 650.84 TROPOJË FIERZË 1,607 0.159 0.04 0.250 0.009 401.53 8,306.66 531.21 TROPOJË LEKBIBAJ 1,205 0.341 0.04 0.217 0.009 262.00 6,301.62 404.98 TROPOJË LLUGAJ 1,787 0.254 0.02 0.236 0.006 421.67 7,119.03 418.34 TROPOJË MARGEGAJ 2,346 0.225 0.04 0.239 0.010 559.86 7,390.66 458.77 TROPOJË TROPOJË 4,117 0.272 0.03 0.231 0.006 952.07 6,875.83 326.83 Portraits of Poverty and Inequality in Albania | Annexes 39 Mean Poverty- Poor District Commune Population s.e. Gini s.e. percapita s.e. Headcount population consumption VLORË BRATAJ 2,835 0.184 0.02 0.243 0.005 688.53 7,928.65 385.09 VLORË HIMARË 2,815 0.116 0.03 0.272 0.006 765.57 9,933.81 336.68 VLORË KOTE 3,522 0.170 0.04 0.240 0.011 844.82 8,021.08 484.83 VLORË NOVOSELË 8,209 0.139 0.05 0.252 0.011 2,070.90 8,751.49 516.21 VLORË ORIKUM 5,457 0.121 0.03 0.259 0.008 1,415.44 9,301.19 361.78 VLORË QENDËR 7,644 0.141 0.05 0.247 0.014 1,889.47 8,608.09 424.54 VLORË SELENICË 2,235 0.178 0.01 0.257 0.003 573.77 8,349.75 186.89 VLORË SEVASTER 1,720 0.177 0.05 0.230 0.013 394.83 7,821.99 556.30 VLORË SHUSHICË 3,981 0.177 0.05 0.248 0.011 986.12 8,108.42 466.42 VLORË VLLAHINË 3,111 0.230 0.03 0.238 0.011 741.19 7,381.93 289.43 VLORË VLORË 78,507 0.124 0.01 0.274 0.003 21,521.72 9,620.53 222.01 VLORË VRANISHT 2,080 0.183 0.03 0.250 0.009 520.32 8,136.27 371.23 VLORË ARMEN 2,965 0.175 0.06 0.238 0.018 706.96 7,970.33 563.34 40 Portraits of Poverty and Inequality in Albania | Annexes Portraits of Poverty and Inequality in Albania | Annexes 41 Results of the 2012 Poverty Maps - New Municipalities Mean per Total Poverty Number of Municipality s.e. Gini s.e. capita s.e. population headcount poor consumption BERAT 59,924 0.129 0.01 0.253 0.01 7,735 9,008 222 URA VAJGURORE 27,286 0.136 0.02 0.241 0.01 3,700 8,608 345 KUÇOVË 31,258 0.124 0.02 0.247 0.01 3,873 8,983 312 SKRAPAR 12,393 0.109 0.03 0.241 0.01 1,345 9,133 491 POLIÇAN 10,922 0.122 0.02 0.237 0.01 1,335 8,752 371 DIBËR 61,619 0.205 0.03 0.076 0.01 12,650 7,232 240 BULQIZË 31,190 0.190 0.03 0.079 0.01 5,926 7,438 291 MAT 27,402 0.153 0.03 0.246 0.01 4,194 8,434 403 KLOS 16,618 0.187 0.04 0.242 0.01 3,112 7,889 412 DURRËS 174,689 0.159 0.01 0.272 0.01 27,773 8,819 174 SHIJAK 27,861 0.152 0.02 0.256 0.01 4,230 8,602 319 KRUJË 59,294 0.135 0.02 0.252 0.01 7,994 8,849 319 ELBASAN 141,224 0.143 0.02 0.257 0.01 20,192 8,818 227 CËRRIK 27,445 0.141 0.03 0.248 0.01 3,881 8,666 370 BELSH 19,503 0.113 0.02 0.245 0.01 2,201 9,077 380 PEQIN 25,399 0.141 0.03 0.247 0.01 3,585 8,637 390 GRAMSH 24,231 0.157 0.03 0.085 0.01 3,815 7,944 292 LIBRAZHD 31,892 0.190 0.03 0.087 0.01 6,060 7,620 277 PRRENJAS 24,905 0.194 0.03 0.081 0.01 4,825 7,437 237 FIER 120,562 0.145 0.02 0.262 0.01 17,475 8,837 193 LUSHNJE 83,644 0.154 0.02 0.253 0.01 12,888 8,470 256 PATOS 22,972 0.154 0.02 0.245 0.01 3,537 8,360 271 ROSKOVEC 21,742 0.192 0.03 0.235 0.01 4,174 7,668 332 DIVJAKË 34,262 0.187 0.03 0.24 0.01 6,418 7,816 288 MALLAKASTËR 27,271 0.166 0.03 0.245 0.01 4,528 8,143 351 GJIROKASTËR 28,545 0.068 0.01 0.27 0.01 1,933 11,215 435 LIBOHOVË 3,667 0.067 0.03 0.255 0.02 246 10,895 888 PËRMET 10,606 0.076 0.02 0.25 0.01 804 10,242 481 KËLCYRË 6,113 0.092 0.03 0.244 0.01 565 9,620 582 TEPELENË 8,889 0.103 0.03 0.256 0.01 917 9,710 526 MEMALIAJ 10,547 0.123 .. .. .. 1,297 9,095 .. DROPULL 3,496 0.078 0.02 0.26 0.01 273 10,670 521 KORÇË 75,292 0.102 0.01 0.267 0.01 7,649 10,061 259 POGRADEC 61,528 0.150 0.02 0.254 0.01 9,222 8,634 274 MALIQ 41,885 0.124 0.02 0.246 0.01 5,185 8,911 310 PUSTEC 3,154 0.050 0.02 0.249 0.01 158 11,213 560 KOLONJË 11,070 0.080 0.02 0.247 0.01 882 10,020 559 DEVOLL 26,720 0.095 0.02 0.257 0.01 2,546 9,821 370 KUKËS 47,963 0.232 0.04 0.079 0.01 11,127 7,067 254 TROPOJË 20,512 0.184 0.03 0.097 0.01 3,769 7,941 356 HAS 16,775 0.233 0.04 0.085 0.01 3,905 7,192 305 LEZHË 64,938 0.162 0.03 0.263 0.01 10,544 8,543 299 MIRDITË 22,088 0.153 0.03 0.251 0.01 3,382 8,503 401 KURBIN 46,249 0.193 0.03 0.243 0.01 8,926 7,770 279 SHKODËR 134,069 0.133 0.01 0.267 0.01 17,850 9,270 210 VAU I DEJËS 30,379 0.143 0.02 0.261 0.01 4,346 8,906 310 42 Portraits of Poverty and Inequality in Albania | Annexes Mean per Total Poverty Number of Municipality s.e. Gini s.e. capita s.e. population headcount poor consumption MALËSI E 30,814 0.149 0.03 0.261 0.01 4,596 8,817 368 MADHE PUKË 11,053 0.168 0.05 0.247 0.01 1,858 8,232 562 FUSHË-ARRËZ 7,391 0.201 0.06 0.245 0.01 1,487 7,791 640 TIRANË 550,163 0.101 .. .. .. 55,561 10,328 .. KAMËZ 104,137 0.237 .. .. .. 24,695 7,454 .. VORË 26,630 0.154 .. .. .. 4,091 8,685 .. KAVAJË 39,846 0.160 0.02 0.259 0.01 6,364 8,580 226 RROGOZHINË 21,739 0.182 0.03 0.245 0.01 3,961 7,975 351 VLORË 103,798 0.129 0.01 0.271 0.01 13,364 9,407 181 HIMARË 7,811 0.140 0.03 0.269 0.01 1,091 9,266 405 SARANDË 20,164 0.112 0.02 0.275 0.01 2,267 9,886 318 KONISPOL 8,236 0.155 0.04 0.242 0.01 1,280 8,299 558 DELVINË 7,598 0.138 0.03 0.256 0.01 1,045 8,937 388 FINIQ 10,529 0.093 0.02 0.249 0.01 978 9,859 570 SELENICË 16,388 0.186 0.04 0.243 0.01 3,045 7,917 405 Note: The indicators for four new municipalities that resulted from merging fractions of former municipalities, as opposed to combinations of former municipalities, were calculated as population weighted averages of the indicators of the former municipalities, and the standard errors cannot be estimated. Moreover, the Gini index cannot be calculated as a weigthed average so it is not available for these municipalities. Portraits of Poverty and Inequality in Albania | Annexes 43 Annex C: Explanation of the Variables and Relevant Data Tables Table C.1 Log Consumption Regression Models with the Mean of the Survey and Census Variables a. Central domain Mean, Mean, Variable Coefficient p value census 2011 LSMS 2012 Intercept 8.56 0.00 Number of children 0 to 5 years 0.08 0.00 0.78 0.79 Number of children 6 to 14 years 0.09 0.00 0.65 0.65 HH has computer 0.18 0.00 0.16 0.18 HH has decoder/TV 0.15 0.00 0.15 0.13 Household size = 1 1.07 0.00 0.06 0.10 Household size = 2 0.71 0.00 0.18 0.19 Household size = 3 0.55 0.00 0.18 0.20 Household size = 4 0.38 0.00 0.24 0.28 Household size = 5 0.24 0.00 0.17 0.14 Household size = 6 0.12 0.01 0.10 0.06 Household head has primary education as highest level 0.16 0.00 0.09 0.10 HH has microwave 0.11 0.00 0.13 0.13 HH do not have central heating −0.28 0.00 0.03 0.04 PSU means Owns livestock or bees 0.24 0.00 0.51 0.49 HH has car 0.21 0.00 0.22 0.22 Highest education level in HH is low −0.30 0.00 0.58 0.58 Household head is female −0.38 0.00 0.13 0.13 Observations 2757 R2 0.46 Adjusted R2 0.46 R2 alpha model 0.00 b. Coastal domain Mean, Mean, Variable Coefficient p value census 2011 LSMS 2012 Intercept_ 8.43 0.00 Number children under age 5=0 0.09 0.00 0.80 0.83 Number children aged 6-14=0 0.08 0.00 0.66 0.68 HH has computer 0.26 0.00 0.15 0.17 Household size = 1 0.96 0.00 0.06 0.07 Household size = 2 0.58 0 0.21 0.21 Household size = 3 0.34 0.00 0.18 0.20 Household size = 4 0.16 0.00 0.24 0.23 Household head has high education 0.08 0.01 0.09 0.09 Household head has low education −0.09 0.00 0.57 0.59 HH has microwave 0.22 0.00 0.15 0.16 Spouse has employment 0.14 0.00 0.10 0.11 PSU means HH has car 0.52 0.00 0.26 0.26 HH has microwave oven −0.29 0.00 0.15 0.15 HH has refrigiator −0.53 0.00 0.94 0.93 MSP_AGE 0.01 0.00 40.34 40.38 HH has TV 0.28 0.01 0.92 0.92 Observations 1837 R2 0.47 Adjusted R2 0.46 R2 alpha model 0.01 44 Portraits of Poverty and Inequality in Albania | Annexes c. Mountain domain Mean, Mean, Variable Coefficient p value census 2011 LSMS 2012 Intercept_ 9.58 0.00 HH has car 0.22 0.00 0.15 0.13 Number children under age 5=0 0.09 0.00 0.71 0.77 Number children aged 6-14=0 0.09 0.00 0.53 0.55 Household size = 1 1.21 0.00 0.03 0.02 Household size = 2 0.77 0.00 0.10 0.11 Household size = 3 0.55 0.00 0.13 0.12 Household size = 4 0.33 0.00 0.21 0.21 Household size = 5 0.20 0.00 0.22 0.24 Household size = 6 0.09 0.07 0.16 0.18 Household head has medium education 0.05 0.03 0.29 0.29 Household head is an employee 0.09 0.00 0.16 0.19 HH has internet 0.18 0.00 0.05 0.07 PSU means HH has fixed telephone line −0.18 0.00 0.14 0.15 Household head is married −1.00 0.00 0.90 0.90 Household head is self-employed 0.19 0.00 0.19 0.19 Highest education in HH is upper 1.07 0.02 0.04 0.03 primary HH own house −0.18 0.00 0.85 0.81 Rooms per capita −0.29 0.00 0.78 0.79 Observations 1128 R2 0.56 Adjusted R2 0.56 d. Tirana domain Mean, Mean, Variable Coefficient p value census 2011 LSMS 2012 Intercept_ 10.69 0.00 Number children aged 6-14=0 −0.07 0.07 0.70 0.68 House is seme-detached house −0.15 0.00 0.11 0.10 Household size −0.41 0.00 3.62 3.79 Household size squared 0.02 0.00 15.68 17.15 Highest education in HH is high 0.18 0.00 0.45 0.44 Highest education in HH is low −0.13 0.00 0.16 0.15 HH has 1 room −0.27 0.00 0.04 0.03 HH has 2 room −0.27 0.00 0.32 0.34 HH has 3 room −0.14 0.00 0.44 0.47 Spouse has employment 0.14 0.00 0.23 0.22 Building built between 1945 and 1960 −0.19 0.01 0.05 0.04 PSU means Deattached house −0.33 0.00 0.66 0.66 Building built between 1961 and 1980 0.31 0.00 0.17 0.17 Observations 648 R2 0.57 Adjusted R2 0.56 R2 alpha model 0.03 Portraits of Poverty and Inequality in Albania | Annexes 45 Table C.2 Intracluster Errors as a Share of Total Errors Coastal Central Mountain Tirana 2 0.040 0.043 0.022 0.030 sˆh 5.930 5.153 5.559 5.013 var( ec• ) 2 ˆ c• ] E [u 5.970 5.195 5.581 5.043 2 0.007 0.008 0.004 0.006 r I (Y | X ) Source: The 2012 Living Standard Measurement Study using calculations following Betti, Neri, and Ballini 2003. Figure C.1 Mean Consumption and Poverty, by Prefecture Tirane 9500 Shkoder Berat Vlore Gjirokaster Korce 9000 Durres 8500 Elbasan Lezhe Fier 8000 Diber Kukes 7500 .1 .15 .2 .25 Figure C.2 Inequality measured by Gini, by prefecture 0.29 0.282 0.285 0.28 0.271 0.271 0.267 0.27 0.262 Percentage (Gini) 0.257 0.259 0.26 0.252 0.253 0.25 0.245 0.24 0.23 0.22 0.211 0.21 0.2 46 Portraits of Poverty and Inequality in Albania | Annexes Figure C.3 Poverty Headcounts and 95 Confidence Intervals, by Prefecture 35% 0.225 30% 0.184 25% 0.154 0.139 20% 0.123 0.165 0.171 15% 0.124 0.111 10% 5% 0.113 0.106 0.127 0% Table C.3 Distribution of Districts, by Region Coastal Central Mountain Tirana Lezhë Devoll Kukës Tirana urban Kurbin Kolonjë Has Tirana, other urban Kavajë Pogradec Tropojë Mallakastër Mirditë Bulqizë Lushnjë Pukë Dibër Delvinë Malësi e Madhe Gramsh Sarandë Mat Librazhd Durrës Kuçovë Fier Skrapar Vlorë Krujë Peqin Gjirokastër Përmet Tepelenë Shkodër Elbasan Berat Korçë Tirana rural Portraits of Poverty and Inequality in Albania | Annexes 47 Main Annex: A closer look - Results by prefecture Berat Prefecture The prefecture of Berat is located in the central Korçë Prefecture to the east. region. It was divided until recently into 20 The poverty rate in the prefecture was 12.6 communes and 5 municipalities. The surface area percent in 2012. The highest poverty rates were of the prefecture is 1,798 square kilometers, and the in the communes of Sinjë (17.8 percent) and population is 141,783. The prefecture is bounded Zhepë (16.9 percent) (map A1). The lowest by Elbasan Prefecture to the north, Fier Prefecture poverty rate occurred in the commune of Çepan to the west, Gjirokastër Prefecture to the south, and (less than 8.0 percent). Map A1: Poverty Rate and Inequality, Berat Prefecture and communes, 2012 a. Poverty rate (headcount) Prefecture of BERAT KOZARE POSHNJË LUMAS KUÇOVË KUTALLI PERONDI OTLLAK URA VAJGURORE ROSHNIK CUKALAT VELABISHT BERAT ZHEPË VELABISHT VERTOP SINJË GJERBËS POLIÇAN BOGOVË TERPAN LESHNJË QENDËR VENDRESHË ÇOROVODË Head Count ratio (%). POTOM < 8.0 8.0 - 12.0 12.1 - 16.0 16.1 - 20.0 ÇEPAN 20.1 - 24.0 24.1 - 28.0 > 28.0 48 Portraits of Poverty and Inequality in Albania | Annexes Prefecture of BERAT Prefecture of BERAT b. Number of the poor KOZARE POSHNJË KUÇOVË LUMAS KOZARE KUTALLI POSHNJË PERONDI KUÇOVË LUMAS KUTALLI OTLLAK URA VAJGURORE PERONDI ROSHNIK CUKALAT VELABISHT OTLLAK URA VAJGURORE BERAT ROSHNIK ZHEPË CUKALAT VELABISHT BERAT VELABISHT VERTOP ZHEPË SINJË VELABISHT VERTOP POLIÇAN GJERBËS SINJË BOGOVË TERPAN POLIÇAN GJERBËS Persons BOGOVË LESHNJË < 300 TERPAN VENDRESHË QENDËR 300 -Persons 600 LESHNJË 601 - 900 ÇOROVODË < 300 901 - 1200 VENDRESHË QENDËR 300 - 600 1201 - 1500 POTOM 601 - 900 ÇOROVODË 1501 - 1800 901 - 1200 1801 - 2100 1201 - 1500 POTOM 2101 - 3000 1501 - 1800 ÇEPAN 1801 - 2100 3001 - 9000 2101 - 3000 ÇEPAN > 9000 3001 - 9000 > 9000 c. The average monthly per capita consumption Per Capita Consumption Per Capita Consumption KOZARE Prefecture of BERAT Prefecture of BERAT POSHNJË LUMAS KUÇOVË KUTALLI PERONDI URA VAJGURORE OTLLAK ROSHNIK CUKALAT VELABISHT BERAT ZHEPË VELABISHT VERTOP SINJË GJERBËS POLIÇAN GJERBËS BOGOVË TERPAN Per capita consumption (ALL/Month) Per capita consumption (ALL/Month) LESHNJË < 7500 LESHNJË < 7500 QENDËR 7500 - 8000 7500 - 8000 8001 - 8500 VENDRESHË 8001 - 8500 ÇOROVODË 8501 - 9000 8501 - 9000 POTOM 9001 - 9500 POTOM 9001 - 9500 9501 - 10000 9501 - 10000 10001 - 10500 10001 - 10500 10501 - 11000 10501 - 11000 11001 - 11500 11001 - 11500 ÇEPAN > 11500 ÇEPAN > 11500 Commune/Municipality Boundary Commune/Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing The boundaries Census and of communes 2011 municipalities have been designed for statistical purpose and haveStandard Living been designed for statistical Measurement Survey purpose – LSMS and 2012 may not reflect exactly the territory of the local units. may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 49 d. Gini coefficient (inequality) Gini Coefficient Gini Coefficient KOZARE Prefecture of BERAT Prefecture of BERAT POSHNJË LUMAS KUÇOVË KUTALLI PERONDI OTLLAK URA VAJGURORE ROSHNIK CUKALAT VELABISHT BERAT ZHEPË VELABISHT VERTOP SINJË GJERBËS POLIÇAN BOGOVË GJERBËS TERPAN LESHNJË Gini index (%) LESHNJË QENDËR Gini index (%) VENDRESHË < 21.0 ÇOROVODË 21.0 - 22.0 < 21.0 POTOM 22.1 - 23.0 21.0 - 22.0 23.1 - 24.0 POTOM 22.1 - 23.0 24.1 - 25.0 23.1 - 24.0 Gini coefficient is a measure of inequality 25.1 - 26.0 100. - 25.0 with values between 0 and 24.1 ÇEPAN > 26.0 25.1 - 26.0 ÇEPAN > 26.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Measurement Living Standard The boundariesSurvey – LSMS and of communes 2012 municipalities may not reflect exactly the territory of the local units. have been designed for statistical purpose and may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The number of the poor in this prefecture was the lowest level of per capita consumption, 17,870 in 2012. The poor were concentrated less than 7,500 ALL. The highest level of per in the municipality of Berat (4,523 people), capita consumption was in the commune of followed by the municipality of Kuçovë (1,547) Çepan (9,854 ALL). and the commune of Kutalli (1,396), while the The Gini coefficient in the prefecture was communes with the lowest number of the poor 23.8 percent. It appeared to be highest in were Çepan (58), Leshnjë (65), and Potom (86). the municipalities of Berat (25.9 percent) The average monthly per capita and Kuçovë (25.2 percent), and lowest in the consumption in the prefecture was 8,785 ALL. communes of Cukalat (22.7 percent), Potom The communes of Sinjë and Tërpan showed (22.2 percent) and Tërpan (22.4 percent). 50 Portraits of Poverty and Inequality in Albania | Annexes Dibër Prefecture The prefecture of Dibër is located in the Republic of Macedonia. the mountain region. It is divided into 4 The poverty rate in the prefecture was 19.5 municipalities and 31 communes. The surface percent in 2012. The highest poverty rate area of the prefecture is 2,586 square kilometers, occurred in the communes of Sllovë (29.4 and the population is 136,829. The prefecture is percent), Zall-Dardhë (28.9 percent), and bounded in the north by Kukës Prefecture, in the Zall-Reç (28.4 percent) (map A2). The lowest west by the prefectures of Durrës and Tirana, in poverty rate was in the municipality of Burrel the northwest by the prefecture of Lezhë, in the (11.2 percent). south by Elbasan Prefecture and in the east by Map A2: Poverty Rate and Inequality, Dibër Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Poverty Head Count ZALL REÇ Prefecture of DIBËR KALA E DODËS Prefecture of DIBËR ZALL DARDHË KALA E DODËS LURË SLLOVË SLLOVË ARRAS FUSHË-ÇIDHËN KASTRIOT ËN ASTRIOT RUKAJ DERJAN MACUKULL FUSHË MUHUR QENDËR TOMIN PESHKOPI MELAN R TOMIN PESHKOPI ULËZ MELAN LUZNI BURREL LIS BAZ SELISHTË MAQELLARË SUÇ MAQELLARË KOMSI SHUPENZË PENZË Head Count ratio (%). GURRË FUSHË BULQIZË Head Count ratio (%). KLOS GJORICË BULQIZË < 8.0 GJORICË 8.0 - 12.0 < 8.0 ZERQAN 12.1 - 16.0 8.0 - 12.0 XIBËR 16.1 - 20.0 12.1 - 16.0 OSTREN 20.1 - 24.0 16.1 - 20.0 MARTANESH 24.1 - 28.0 OSTREN 20.1 - 24.0 > 28.0 24.1 - 28.0 TREBISHT > 28.0 TREBISHT Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard The boundaries Measurement of communes Survey – LSMS 2012 and municipalities may not reflect exactly the territory of the local units. have been designed for statistical purpose and may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 51 b. Number of the poor Number of poor individuals ZALL REÇ Number of poor individuals Prefecture of DIBËR KALA E DODËS ZALL DARDHË Prefecture of DIBËR LURË SLLOVË KALA E DODËS ARRAS SLLOVË FUSHË-ÇIDHËN KASTRIOT KASTRIOT RUKAJ MACUKULL FUSHË MUHUR PESHKOPI DERJAN MELAN QENDËR TOMIN ULËZ PESHKOPI LUZNI MELAN BURREL OMIN LIS BAZ SELISHTË MAQELLARË Persons SUÇ KOMSI SHUPENZË < 300 MAQELLARË 300 - 600 Persons NZË GURRË FUSHË BULQIZË 601 - 900 < 300 KLOS 901 - 1200 300 - 600 GJORICË BULQIZË 1201 - 1500 601 - 900 1501 - 1800 901 - 1200 ZERQAN GJORICË 1801 - 2100 1201 - 1500 XIBËR 2101 - 3000 1501 - 1800 1801 - 2100 3001 - 9000 OSTREN MARTANESH 2101 - 3000 > 9000 3001 - 9000 OSTREN > 9000 TREBISHT Source: The boundaries of communes and municipalities TREBISHT Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The boundaries of communes and municipalities have been designed for statistical purpose and may not reflect exactly the territory of the local units. c. The average monthly per capita consumption Per Capita Consumption Per Capita Consumption ZALL REÇ Prefecture of DIBËR KALA E DODËS Prefecture of DIBËR ZALL DARDHË LURË KALA E DODËS SLLOVË ARRAS FUSHË-ÇIDHËN SLLOVË KASTRIOT OT RUKAJ MACUKULL DERJAN FUSHË MUHUR QENDËR TOMIN PESHKOPI MELAN ULËZ MINPESHKOPI LUZNI MELAN BURREL LIS BAZ SELISHTË MAQELLARË SUÇ KOMSI SHUPENZË MAQELLARË Per capita consumption (ALL/Month) FUSHË BULQIZË < 7500 ZË GURRË Per capita consumption (ALL/Month) 7500 - 8000 KLOS BULQIZË GJORICË 8001 - 8500 < 7500 8501 - 9000 7500 - 8000 ZERQAN 9001 - 9500 GJORICË 8001 - 8500 XIBËR 9501 - 10000 8501 - 9000 10001 - 10500 9001 - 9500 OSTREN MARTANESH 10501 - 11000 9501 - 10000 11001 - 11500 10001 - 10500 OSTREN > 11500 10501 - 11000 TREBISHT Commune/Municipality Boundary 11001 - 11500 > 11500 Source: TREBISHT The boundaries of communes and municipalities Population and Housing Commune/Municipality Census 2011 Boundary have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The boundaries of communes and municipalities have been designed for statistical purpose and may not reflect exactly the territory of the local units. 52 Portraits of Poverty and Inequality in Albania | Annexes d. Gini coefficient (inequality) Gini Coefficient Gini Coefficient ZALL REÇ Prefecture of DIBËR KALA E DODËS Prefecture of DIBËR ZALL DARDHË LURË SLLOVË KALA E DODËS ARRAS FUSHË-ÇIDHËN SLLOVË KASTRIOT N STRIOT RUKAJ DERJAN MACUKULL FUSHË MUHUR QENDËR TOMIN PESHKOPI MELAN ULËZ TOMIN PESHKOPI LUZNI MELAN BURREL LIS BAZ SELISHTË MAQELLARË SUÇ KOMSI SHUPENZË MAQELLARË GURRË FUSHË BULQIZË PENZË KLOS Gini index (%) GJORICË BULQIZË < 21.0 ZERQAN Gini index (%) 21.0 - 22.0 GJORICË XIBËR 22.1 - 23.0 < 21.0 23.1 - 24.0 21.0 - 22.0 OSTREN MARTANESH 24.1 - 25.0 22.1 - 23.0 25.1 - 26.0 23.1 Gini - 24.0 is a measure of inequality coefficient OSTREN > 26.0 with 24.1values - 25.0between 0 and 100. TREBISHT 25.1 - 26.0 > 26.0 TREBISHT Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The boundaries of communes and municipalities have been designed for statistical purpose and may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The number of the poor in the prefecture was (9,101 ALL). The communes of Kala e Dodës, estimated at 26,731. The largest number of poor Luzni, Selishtë, Sllovë, Zall-Dardhë and Zall- was in the municipality of Peshkopi (2,745 poor Reç had the lowest per capita consumption, less people) and the commune of Maqellarë (2,561). than 7,000 ALL. The inequality measure in the The average monthly per capita consumption in prefecture was 23.1 percent. The highest level of this prefecture was 7,551 ALL, which is below inequality was in the commune of Zall-Dardhë the national average. The highest per capita (25.2 percent) and the lowest level of inequality consumption was in the municipality of Burrel in commune of Kala e Dodes (20.4 percent) Portraits of Poverty and Inequality in Albania | Annexes 53 Durrës Prefecture The prefecture of Durrës is located in the prefecture of Tirana. coastal region. It is divided into 6 municipalities The poverty rate in the prefecture was 15.3 and 10 communes. The surface area of the percent in 2012. The highest poverty rate was in prefecture is 766 square kilometers and the the municipalities of Sukth (21.1 percent) and population is 261,844. The prefecture is boun­ Manëz (20.9 percent), followed by the commune ded on the east by the prefecture of Dibër, on of Katund i Ri (20.2 percent) (map A3). The the north by the prefecture of Lezhë, on the west lowest poverty rate occurred in the commune of by the Adriatic Sea, and on the south by the Bubq (9.0 percent). Poverty Head Count Prefecture of DURRËS Map A3: Poverty Rate and Inequality, Durrës Prefecture and communes, 2012 Poverty Head Count a. Poverty rate (headcount) Prefecture of DURRËS ISHËM KODËR THUMANË CUDHI KRUJË BUBQ ISHËM KODËR THUMANË FUSHË KRUJË CUDHI KRUJË NIKËL MANËZ SUKTH BUBQ FUSHË KRUJË KATUND I RI NIKËL MANËZ SUKTH MAMINAS XHAFZOTAJ SHIJAK KATUNDDURRËS I RI GJEPALAJ MAMINAS RRASHBULL XHAFZOTAJ Head Count ratio in (%). SHIJAK DURRËS < 8.0 GJEPALAJ 8.0 - 12.0 12.1 - 16.0 RRASHBULL 16.1 - 20.0 Head Count ratio in (%). 20.1 - 24.0 24.1 - 28.0 < 8.0 8.0 - 12.0 > 28.0 12.1 - 16.0 16.1 - 20.0 Source: The boundaries of communes and municipalities 20.1 - 24.0 Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 24.1 - 28.0 may not reflect exactly the territory of the local units. > 28.0 54 Portraits of Poverty and Inequality in Albania | Annexes Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Prefecture of DURRËS KODËR THUMANË ISHËM CUDHI KRUJË b. Number of the poor BUBQ FUSHË KRUJË KODËR THUMANË ISHËM MANËZ NIKËL SUKTH CUDHI KRUJË KATUND I RI BUBQ FUSHË KRUJË MAMINAS MANËZ NIKËL SUKTH XHAFZOTAJ DURRËS SHIJAK Persons RRASHBULL GJEPALAJ KATUND I RI < 300 Per Capita Consumption 300 - 600 MAMINAS 601 - 900 901 - 1200 XHAFZOTAJ Prefecture of DURRËS 1201 - 1500 DURRËS SHIJAK 1501 - 1800 Persons 1801 - 2100 RRASHBULL GJEPALAJ 2101 - 3000 < 300 Per Capita Consumption 300 - 600 3001 - 9000 601 - 900 901 - 1200 > 9000 1201of Prefecture DURRËS - 1500 1501 - 1800 KODËR THUMANË 1801 - 2100 Source: ISHËM The boundaries of communes and municipalities c. The average monthly per capita Population consumption and Housing Census 2011 2101 - 3000 have been designed CUDHI for statistical purpose and Living Standard Measurement Survey – LSMS 2012 KRUJË may not reflect exactly the territory of the local units. 3001 - 9000 BUBQ > 9000 FUSHË KRUJË ISHËM KODËR THUMANË NIKËL Source: MANËZ The boundaries of communes and municipalities SUKTH Population and Housing Census 2011 CUDHI purpose and have been designed for statistical Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. KRUJË KATUND I RI BUBQ FUSHË KRUJË MAMINAS NIKËL MANËZ SUKTH XHAFZOTAJ SHIJAK DURRËS GJEPALAJ KATUND I RI Per capita consumption (ALL/Month) RRASHBULL MAMINAS < 7500 7500 - 8000 8001 - 8500 XHAFZOTAJ SHIJAK 8501 - 9000 DURRËS 9001 - 9500 9501 - 10000 GJEPALAJ 10001 - 10500 (ALL/Month) Per capita consumption 10501 - 11000 RRASHBULL < 7500 11001 - 11500 7500 - 8000 > 11500 8001 - 8500 Commune/Municipality Boundary 8501 - 9000 Source: The boundaries of communes and municipalities Population and Housing Census 2011 9001 have 9500 designed for statistical purpose and - been Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 9501 - 10000 10001 - 10500 Portraits of Poverty and Inequality in Albania | Annexes 10501 - 11000 55 11001 - 11500 > 11500 Commune/Municipality Boundary Gini Coefficient Prefecture of DURRËS ISHËM KODËR THUMANË d. Gini coefficient (inequality) CUDHI KRUJË BUBQ FUSHË KRUJË ISHËM KODËR THUMANË CUDHI KRUJË NIKËL MANËZ SUKTH BUBQ FUSHË KRUJË KATUND I RI NIKËL MANËZ MAMINAS SUKTH XHAFZOTAJ SHIJAK KATUND I RI DURRËS MAMINAS GJEPALAJ RRASHBULL XHAFZOTAJ DURRËS SHIJAK Gini index (%) < 21.0 GJEPALAJ 21.0 - 22.0 22.1 - 23.0 RRASHBULL 23.1 - 24.0 Gini index (%) 24.1 - 25.0 < 21.0 25.1 - 26.0 Gini coefficient is a measure of inequality 21.0 - 22.0 > 26.0 with values between 0 and 100. 22.1 - 23.0 23.1 - 24.0 24.1 - 25.0 Source: The boundaries of communes and municipalities 25.1 - 26.0 Population and Housing Census 2011 have been designed for statistical purpose and Gini coefficient is a measure of inequality > 26.0 Living Standard Measurement Survey – LSMS 2012 with values between 0 and 100. may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The number of poor in the prefecture was and Durrës (9,351 ALL). The lowest level of 39,940 in 2012. The highest number of poor was consumption was in the municipalities of Sukth in the municipality of Durrës, around 16,000 (7,514 ALL), Katund i Ri (7,607 ALL), Gjepalaj individuals, with a poverty rate of 14.1 percent. (7,659 ALL) and Manëz (7,787 ALL). The average monthly per capita consumption The level of inequality in the prefecture was in the prefecture was almost the same as the 24.9 percent. The highest level of inequality was average per capita consumption in the country, in the municipalities of Durrës (27.9 percent) 8,481 ALL. The highest per capita consumption and Shijak (26.6 percent). The lowest poverty was in the commune of Bubq (9,788 ALL) and rate occurred in the commune of Gjepalaj (23.6 in the municipalities of Shijak (9,382 ALL) percent). 56 Portraits of Poverty and Inequality in Albania | Annexes Elbasan Prefecture The prefecture of Elbasan is located partly in The prefecture of Elbasan had a poverty rate the mountain region and partly in the central of 15.2 percent in 2012. Communes that are region. It is divided into 7 municipalities and 43 in the mountain region had a higher poverty communes. The surface area of the prefecture is rate compared with the other communes in 3,278 square kilometers, and the population is the prefecture. The highest poverty rates were 294,599. The prefecture is bounded on the north in the communes of Orenjë (30.4 percent), by the prefecture of Dibër, on the northwest Poroçan (26.8 percent) and Lunik (24.5 percent) by the prefecture of Tirana, in the west by the (map A4). The lowest poverty rates were in the prefecture of Fier, in the south by the prefecture municipality of Librazhd and the communes of Berat, in the southeast by the prefecture of of Belsh, Gjergjan, Gostimë, Kushovë, and Korçë, and in the east by Republic of Macedonia. Shirgjan, where the rates fell below 12.0 percent. Map A4: Poverty Rate and Inequality, Elbasan Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Poverty Head Count STËBLEVË Prefecture of ELBASAN ËBLEVË Prefecture of ELBASAN LUNIK ORENJË LABINOT MAL FUNARË QENDËR LIBRAZHD GRACEN LABINOT FUSHË HOTOLISHT OTOLISHT BRADASHESH POLIS ELBASAN PAPËR SHUSHICË PAJOVË PËRPARIM QUKËS RRAJCË KARINË SHIRGJAN QUKËS RRAJCË PËRRENJAS PEQIN PËRRENJAS GJOCAJ CËRRIK GJERGJAN STRAVAJ SHALËS GJINAR SHEZË GOSTIMË TREGAN AJ ZAVALIN BELSH RRASË KLOS POROÇAN FIERZË Head Count ratio (%). KAJAN MOLLAS Head Count ratio (%). SULT PISHAJ GREKAN < 8.0 < 8.0 GRAMSH 8.0 - 12.0 8.0 - 12.0 KUKUR 12.1 - 16.0 12.1 - 16.0 TUNJË KODOVJAT 16.1 - 20.0 16.1 - 20.0 20.1 - 24.0 20.1 - 24.0 LENIE 24.1 - 28.0 KUSHOVË 24.1 - 28.0 > 28.0 SKËNDERBEGAS > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and The boundaries of communes and municipalities Living Standard have been designed purposeSurvey Measurement for statistical and – LSMS 2012 may not reflect exactly the territory of the local units. may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 57 b. Number of the poor Number of poor indi Number of poor individuals STËBLEVË Prefecture of ELBAS Prefecture of ELBASAN ORENJË LUNIK FUNARË LABINOT MAL QENDËR GRACEN HOTOLISHT LIBRAZHD BRADASHESH LABINOT FUSHË POLIS ELBASAN SHUSHICË PAPËR QUKËS PËRPARIM PAJOVË RRAJCË KARINË SHIRGJAN PËRRENJAS KËS RRAJCË PEQIN PËRRENJAS GJERGJAN GJOCAJ CËRRIK SHEZË SHALËS TREGAN GJINAR Persons STRAVAJ GOSTIMË Persons ZAVALIN < 300 BELSH RRASË KLOS 300 - 600 < 300 601 - 900 300 - 600 901 - 1200 601 - 900 MOLLAS POROÇAN FIERZË KAJAN GREKAN PISHAJ 1201 - 1500 901 - 1200 SULT 1501 - 1800 1201 - 1500 1801 - 2100 1501 - 1800 GRAMSH KUKUR 1801 - 2100 TUNJË 2101 - 3000 KODOVJAT 2101 - 3000 3001 - 9000 KUSHOVË LENIE 3001 - 9000 SKËNDERBEGAS > 9000 > 9000 Source: The boundaries of communes and muni Population and Housing Census 2011 have been designed for statistical purpo Standard Living and The boundaries of communes Measurement Survey – LSMS 2012 municipalities may not reflect exactly the territory of the have been designed for statistical purpose and may not reflect exactly the territory of the local units. c. The average monthly per capita consumption Per Capita Consumption Per Capita Consumption Prefecture of ELBASAN Prefecture of ELBASA STËBLEVË LUNIK ORENJË LABINOT MAL FUNARË QENDËR LIBRAZHD GRACEN LABINOT FUSHË HOTOLISHT BRADASHESH POLIS ELBASAN PAPËR SHUSHICË QUKËS PAJOVË RRAJCË PËRPARIM QUKËS RRAJCË KARINË SHIRGJAN PËRRENJAS PËRRENJAS PEQIN GJOCAJ CËRRIK GJERGJAN SHEZË STRAVAJ SHALËS GJINAR GOSTIMË TREGAN ZAVALIN BELSH Per capita consumption (ALL/Month) RRASË KLOS Per capita consumption (ALL/Month) < 7500 < 7500 7500 - 8000 POROÇAN 7500 - 8000 FIERZË MOLLAS KAJAN SULT PISHAJ 8001 - 8500 GREKAN 8001 - 8500 8501 - 9000 8501 - 9000 GRAMSH KUKUR 9001 - 9500 9001 - 9500 TUNJË 9501 - 10000 KODOVJAT 9501 - 10000 10001 - 10500 10001 - 10500 10501 - 11000 KUSHOVË LENIE 10501 - 11000 11001 - 11500 SKËNDERBEGAS 11001 - 11500 > 11500 > 11500 Commune/Municipality Boundary Commune/Municipality Boundary Source: The boundaries of communes and municipalities The boundaries of communes and municipali Population have been designed for and Housing statistical Census purpose 2011 and have been designed for statistical purpose an may not reflect exactly Living the territory Standard of the local Measurement units. Survey – LSMS 2012 may not reflect exactly the territory of the loca 58 Portraits of Poverty and Inequality in Albania | Annexes d. Gini coefficient (inequality) Gini Coefficient Gini Coefficient Prefecture of ELBASAN Prefecture of ELBASAN STËBLEVË STËBLEVË LUNIK ORENJË LABINOT MAL FUNARË QENDËR LIBRAZHD GRACEN LABINOT FUSHË HOTOLISHT HOTOLISHT BRADASHESH POLIS ELBASAN PAPËR SHUSHICË PAJOVË RRAJCË PËRPARIM QUKËS KARINË QUKËS RRAJCË SHIRGJAN PËRRENJAS PËRRENJAS PEQIN GJOCAJ CËRRIK GJERGJAN STRAVAJ SHEZË SHALËS GJINAR RAVAJ GOSTIMË TREGAN ZAVALIN BELSH RRASË KLOS POROÇAN FIERZË MOLLAS Gini index (%) KAJAN Gini index (%) GREKAN SULT PISHAJ < 21.0 < 21.0 GRAMSH KUKUR 21.0 - 22.0 21.0 - 22.0 TUNJË 22.1 - 23.0 KODOVJAT 22.1 - 23.0 23.1 - 24.0 23.1 - 24.0 24.1 - 25.0 LENIE 24.1 - 25.0 KUSHOVË NIE 25.1 - 26.0 25.1 is Gini coefficient a measure of inequality - 26.0 SKËNDERBEGAS with values between 0 and 100. > 26.0 > 26.0 Source: The boundaries of communes and municipalities Population The Housing and of boundaries Census communes and 2011 municipalities have been designed for statistical purpose and Living have Standard been Measurement designed Survey for statistical – LSMS purpose and 2012 may not reflect exactly the territory of the local units. may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The number of the poor in the prefecture was the commune of Poroçan (6,868 ALL). The 44,727 in 2012. About 22.0 percent of the poor municipalities of Belsh, Elbasan, and Librazhd in the prefecture were living in the municipality and the communes of Gostimë and Shirgjan of Elbasan, which had the highest density of had consumption levels above 9,000 ALL, poor compared with other municipalities and significantly above the average across the communes. The areas with the lowest number prefecture. of poor were the communes of Kushovë and The measure of inequality in the prefecture Stëblevë. was 23.7 percent. The highest Gini coefficient The average monthly per capita consumption was in the municipalities of Elbasan and Peqin in the prefecture was 8,192 ALL. The poorest (above 26 percent). The lowest level of inequality commune, Orenjë, also exhibited the lowest were in commune of Hotolisht, Lenie and Sult consumption level (6,562 ALL), followed by (less than 22 percent). Portraits of Poverty and Inequality in Albania | Annexes 59 Fier Prefecture The prefecture of Fier is located in the coastal prefecture of Vlorë. region. It is divided into 6 municipalities and 36 The prefecture of Fier had a poverty rate of communes. The surface area of the prefecture 15.7 percent in 2012. The lowest poverty rate is 1,887 square kilometers, and the population was in the municipalities of Fier and Ballsh, is 310,453. The prefecture is bounded on the respectively, 11.0 percent and 11.4 percent north by the prefecture of Durrës, on the (map A5). The highest poverty rates were in the northwest by the Adriatic Sea, on the northeast communes of Selitë (22.1 percent) and Tërbuf by the prefecture of Elbasan, on the east by the (21.6 percent). prefecture of Berat and on the south by the Poverty Head Count Map A5:Prefecture Poverty FIER and Inequality, Fier Prefecture and communes, 2012 ofRate TËRBUF a. Poverty rate (headcount) DUSHK DIVJAKË BALLAGAT Poverty Head Count GOLEM HYSGJOKAJ GRABIAN Prefecture of FIER KARBUNARË LUSHNJE TËRBUF DUSHK GRADISHTË FIERSHEGAN DIVJAKË BALLAGAT KRUTJE KOLONJË ALLKAJ GOLEM HYSGJOKAJ OFSHË GRABIAN BUBULLIMË KARBUNARË LUSHNJE QENDËR MBROSTAR STRUM RREMAS ROSKOVEC GRADISHTË FIERSHEGAN KRUTJE FIER KOLONJË ALLKAJ ZHARRËS KUMAN LIBOFSHË KURJAN BUBULLIMË PATOS TOPOJË PORTËZ RUZHDIE QENDËR MBROSTAR STRUM NGRAÇAN ROSKOVEC ULL RUZHDIE QENDËR DERMENAS ARANITAS FIER ZHARRËS CAKRAN KUMAN BALLSH KURJAN LEVAN PATOS GRESHICË HEKAL PORTËZ RUZHDIE SELITË NGRAÇAN FRAKULL RUZHDIE FRATAR QENDËR ARANITAS CAKRAN BALLSH KUTË Head Count ratio (%). GRESHICË HEKAL < 8.0 SELITË 8.0 - 12.0 12.1 - 16.0 FRATAR 16.1 - 20.0 20.1 - 24.0 24.1 - 28.0 KUTË > 28.0 Head Count ratio (%). < 8.0 The boundaries of communes and municipalities have been designed for statistical purpose and 8.0 - 12.0 012 may not reflect exactly the territory of the local units. 12.1 - 16.0 16.1 - 20.0 20.1 - 24.0 24.1 - 28.0 > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 60 Portraits of Poverty and Inequality in Albania | Annexes BALLAGAT HYSGJOKAJ GRABIAN GOLEM LUSHNJE KARBUNARË GRADISHTË FIERSHEGAN Number of poor individuals Prefecture of FIER KRUTJE KOLONJË ALLKAJ b. Number of the poor TËRBUF BUBULLIMË DUSHK DIVJAKË MBROSTAR BALLAGAT STRUM ROSKOVEC HYSGJOKAJ GRABIAN GOLEM R ZHARRËS KUMAN LUSHNJE KARBUNARË KURJAN PATOS RREMAS RUZHDIE GRADISHTË FIERSHEGAN PORTËZ NGRAÇAN KRUTJE KOLONJË ALLKAJ LIBOFSHË QENDËR CAKRAN ARANITAS BUBULLIMË BALLSH TOPOJË QENDËR MBROSTAR GRESHICË STRUM HEKAL ROSKOVEC SELITË DERMENAS Persons FIER ZHARRËS KUMAN FRATAR < 300 300 - 600 KURJAN PATOS LEVAN 601 - 900 KUTË 901 - 1200 PORTËZ RUZHDIE 1201 - 1500 NGRAÇAN FRAKULL 1501 - 1800 QENDËR 1801 - 2100 CAKRAN ARANITAS 2101 - 3000 BALLSH 3001 - 9000 GRESHICË HEKAL Per Capita Consumption SELITË > 9000 Persons Prefecture of FIER FRATAR < 300 The boundaries of communes and municipalities TËRBUF have been designed for statistical purpose and 300 - 600 may not reflect exactly the territory of the local units. 601 - 900 KUTË DUSHK 901 - 1200 BALLAGAT 1201 - 1500 1501 - 1800 GOLEM HYSGJOKAJ GRABIAN 1801 - 2100 c. The average monthly per capita consumption 2101 - 3000 LUSHNJE KARBUNARË Per Capita Consumption 3001 - 9000 GRADISHTË Prefecture of FIER FIERSHEGAN TËRBUF > 9000 KRUTJE KOLONJË ALLKAJ DUSHK Source: The boundaries of communes and municipalities Population and Housing Census 2011 DIVJAKË BALLAGAT have been designed for statistical purpose and BUBULLIMË Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. GOLEM HYSGJOKAJ GRABIAN MBROSTAR STRUM ROSKOVEC KARBUNARË LUSHNJE R ER ZHARRËS KUMAN RREMAS GRADISHTË FIERSHEGAN KURJAN PATOS KRUTJE KOLONJË ALLKAJ PORTËZ RUZHDIE LIBOFSHË BUBULLIMË NGRAÇAN RUZHDIE TOPOJË QENDËR ARANITAS CAKRAN MBROSTAR STRUM BALLSH ROSKOVEC QENDËR GRESHICË DERMENAS HEKAL FIER ZHARRËS KUMAN SELITË KURJAN LEVAN PATOS FRATAR PORTËZ RUZHDIE < 7500 NGRAÇAN KUTË 7500 - 8000 FRAKULL RUZHDIE 8001 - 8500 QENDËR ARANITAS 8501 - 9000 CAKRAN BALLSH 9001 - 9500 9501 - 10000 GRESHICË HEKAL 10001 - 10500 SELITË 10501 - 11000 11001 - 11500 FRATAR > 11500 Commune/Municipality Boundary < 7500 KUTË 7500 - 8000 The boundaries of communes and municipalities have been designed for statistical purpose and 8001 - 8500 may not reflect exactly the territory of the local units. 8501 - 9000 9001 - 9500 9501 - 10000 10001 - 10500 Portraits of Poverty and Inequality in Albania | Annexes 10501 - 11000 61 11001 - 11500 > 11500 Commune/Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Prefecture of FIER TËRBUF DUSHK Gini Coefficient KË BALLAGAT GOLEM Prefecture of FIER GRABIAN HYSGJOKAJ d. Gini coefficient (inequality) TËRBUF KARBUNARË LUSHNJE DUSHK GRADISHTË DIVJAKË BALLAGAT FIERSHEGAN GOLEM HYSGJOKAJ KRUTJE KOLONJË ALLKAJ GRABIAN Ë BUBULLIMË KARBUNARË LUSHNJE RREMAS MBROSTAR GRADISHTË STRUM FIERSHEGAN ROSKOVEC QENDËR KRUTJE KOLONJË ALLKAJ FIER LIBOFSHË ZHARRËS KUMAN BUBULLIMË KURJAN TOPOJË PATOS PORTËZ RUZHDIE MBROSTAR STRUM ROSKOVEC NGRAÇAN QENDËR RUZHDIE DERMENAS QENDËR FIER ZHARRËS ARANITAS KUMAN CAKRAN BALLSH KURJAN LEVAN PATOS GRESHICË PORTËZ RUZHDIE HEKAL SELITË NGRAÇAN FRAKULL RUZHDIE FRATAR QENDËR ARANITAS CAKRAN BALLSH Gini index (%) KUTË < 21.0 GRESHICË HEKAL 21.0 - 22.0 SELITË 22.1 - 23.0 23.1 - 24.0 FRATAR 24.1 - 25.0 25.1 - 26.0 Gini index (%) > 26.0 KUTË < 21.0 The boundaries of communes and municipalities 21.0 - 22.0 have been designed for statistical purpose and 12 may not reflect exactly the territory of the local units. 22.1 - 23.0 23.1 - 24.0 24.1 - 25.0 Gini coefficient is a measure of inequality with values between 0 and 100. 25.1 - 26.0 > 26.0 Census 2011 Source: Population and HousingSource: The boundaries of communes and municipalities have The boundaries of communes andbeen designed for municipalities Population and Housing Census 2011 Living Standard Measurement Survey – LSMS Living Standard 2012 Measurement Survey – LSMS 2012 may not reflectthe statistical purpose and may not reflect exactly exactlyterritory oflocal the territory of the the have been designed for statistical purpose and local unit units. About 12.5 percent of the poor in the was above 9,000 ALL. The communes of Strum prefecture of Fier were living in the municipality and Tërbuf had the lowest levels of average of Fier, followed by about 8.0 percent in the consumption per capita in the prefecture, municipality of Lushnjë. The commune of respectively, 7,287 ALL and 7,390 ALL. Selitë had one of the lowest numbers of poor The level of inequality in the prefecture is individuals and the highest poverty rates in the below the average of the country, 23.6 percent. prefecture. The highest inequality measure was in the The average monthly per capita consumption municipality of Fier (27.1 percent), followed in the prefecture was 7,979 ALL. The highest by the municipality of Lushnjë (26.5 percent). level of consumption was in the municipalities The lowest level of inequality is in commune of of Ballsh, Fier, and Lushnjë, where the level Ngraçan (20.7 percent). 62 Portraits of Poverty and Inequality in Albania | Annexes Gjirokastër Prefecture The prefecture of Gjirokastër is located in the Albania in 2012 relative to the other prefectures central region. It is divided into 6 municipalities and to the national average (8.3 percent). The and 26 communes. The surface area is 2,884 highest poverty rates were in the municipalities square kilometers, and the population is 71,863. of Krahës (14.3 percent), Qesarat (13.6 percent), The prefecture is bounded on the north by the Lopës (13.3 percent), and Kurvelesh (12.9 prefecture of Berat, on the northwest by the percent) (map A6). The communes with the prefecture of Fier, on the south and west by the lowest poverty rates were Zagori, Pogon, and prefecture of Vlorë, on the southeast by Greece, Odrie at, respectively, with 2.6 percent, 3.1 and on the east by the prefecture of Korçë. percent, and 3.9 percent. The prefecture had the lowest poverty rate in Poverty Head Count Prefecture of GJIROKASTËR Map A6: Poverty Rate and Inequality, Gjirokastër Prefecture and communes, 2012 BUZ a. Poverty rate (headcount) Poverty Head Count BALLABAN Prefecture of GJIROKASTËR LUFTINJË SUKË FRASHËR MEMALIAJFSHAT MEMALIAJ DISHNICË BUZ TEPELENË QENDËR KRAHËS BALLABAN KËLCYRË QENDËR QESARAT LUFTINJË PËRMET PETRAN SUKË ODRIE FRASHËR LOPËS MEMALIAJFSHAT MEMALIAJ ZAGORI DISHNICË PICAR TEPELENË QENDËR LUNXHËRI ÇARÇOVË KËLCYRË QENDËR CEPO ANTIGONË QENDËR LIBOHOVË KURVELESH GJIROKASTËR PËRMET POGON PETRAN LAZARAT ODRIE LIBOHOVË ZAGORI PICAR DROPULL I POSHTËM LUNXHËRI ÇARÇOVË CEPO ANTIGONË QENDËR LIBOHOVË GJIROKASTËR POGON LAZARAT LIBOHOVË DROPULL I SIPËRM DROPULL I POSHTËM Head Count ratio in (%). < 8.0 8.0 - 12.0 12.1 - 16.0 16.1 - 20.0 DROPULL I SIPËRM 20.1 - 24.0 24.1 - 28.0 > 28.0 Head Count ratio in (%). < 8.0 The boundaries of communes and municipalities have been designed for statistical purpose and 8.0 - 12.0 – LSMS 2012 may not reflect exactly the territory of the local units. 12.1 - 16.0 16.1 - 20.0 20.1 - 24.0 24.1 - 28.0 > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Portraits of Poverty and Living Standard Inequality Measurement 2012 Albania | Annexes Survey – LSMS in may not reflect exactly the territory of the local units. 63 Number of poor individuals BALLABAN SUKË Prefecture of GJIROKASTËR FRASHËR b. Number of the poor DISHNICË BUZ QENDËR KRAHËS BALLABAN KËLCYRË QESARAT LUFTINJË PËRMET PETRAN SUKË FRASHËR FSHAT MEMALIAJ ODRIE LOPËS ZAGORI DISHNICË MEMALIAJ TEPELENË QENDËR LUNXHËRI ÇARÇOVË KËLCYRË QENDËR ANTIGONË GJIROKASTËR KURVELESH QENDËR LIBOHOVË POGON PËRMET PETRAN LAZARAT ODRIE LIBOHOVË ZAGORI PICAR DROPULL I POSHTËM LUNXHËRI ÇARÇOVË CEPO ANTIGONË GJIROKASTËR Persons QENDËR LIBOHOVË POGON < 300 LAZARAT 300 - 600 601 Consumption Per Capita LIBOHOVË DROPULL I SIPËRM - 900 901 - 1200 1201 -of Prefecture 1500GJIROKASTËR DROPULL I POSHTËM 1501 - 1800 1801 - 2100 Persons 2101 - 3000 < 300 3001 - 9000 300 - 600 DROPULL I SIPËRM BALLABAN 601 - 900 > 9000 Per Capita 901 - 1200 Consumption 1201 - 1500 SUKË Prefecture 1501 - 1800 of GJIROKASTËR c. FRASHËR The average of communes and per monthly The boundaries capita consumption municipalities 1801 - 2100 have been designed for statistical purpose and DISHNICË may not reflect exactly the territory of the local units. 2101 - 3000 QENDËR 3001 - 9000 BUZ KËLCYRË KRAHËS > 9000 BALLABAN PËRMET PETRAN QESARAT LUFTINJË ODRIE Source: The boundaries of communes and municipalities SUKË ZAGORI Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 FSHAT MEMALIAJ FRASHËR may not reflect exactly the territory of the local units. LOPËS MEMALIAJ DISHNICË LUNXHËRI ÇARÇOVË TEPELENË QENDËR ANTIGONË QENDËR LIBOHOVË KËLCYRË GJIROKASTËR POGON QENDËR KURVELESH LAZARAT PËRMET PETRAN LIBOHOVË ODRIE ZAGORI DROPULL I POSHTËM PICAR LUNXHËRI ÇARÇOVË CEPO ANTIGONË QENDËR LIBOHOVË GJIROKASTËR POGON DROPULL I SIPËRM Per capita consumption (ALL/Month) LAZARAT < 7500 LIBOHOVË 7500 - 8000 8001 - 8500 DROPULL I POSHTËM 8501 - 9000 9001 - 9500 9501 - 10000 10001 - 10500 10501 - 11000 11001 - 11500 DROPULL I SIPËRM Per capita consumption (ALL/Month) > 11500 < 7500 Commune/Municipality Boundary 7500 - 8000 8001 - 8500 The boundaries of communes and municipalities have been designed for statistical purpose and 8501 - 9000 2 may not reflect exactly the territory of the local units. 9001 - 9500 9501 - 10000 64 in -Albania Portraits of Poverty and Inequality 10001 10500 10501 - 11000 | Annexes 11001 - 11500 > 11500 Commune/Municipality Boundary Prefecture of GJIROKASTËR Gini Coefficient BALLABAN d. Gini coefficient (inequality) Prefecture of GJIROKASTËR SUKË FRASHËR IAJ DISHNICË BUZ QENDËR KRAHËS BALLABAN KËLCYRË DËR QESARAT LUFTINJË PËRMET PETRAN SUKË ODRIE FRASHËR LOPËS MEMALIAJFSHAT MEMALIAJ ZAGORI DISHNICË TEPELENË QENDËR LUNXHËRI ÇARÇOVË KËLCYRË QENDËR ANTIGONË QENDËR LIBOHOVË KURVELESH GJIROKASTËR PËRMET POGON PETRAN LAZARAT ODRIE LIBOHOVË ZAGORI PICAR DROPULL I POSHTËM LUNXHËRI ÇARÇOVË CEPO ANTIGONË QENDËR LIBOHOVË GJIROKASTËR POGON LAZARAT LIBOHOVË DROPULL I SIPËRM Gini index (%) DROPULL I POSHTËM < 21.0 21.0 - 22.0 22.1 - 23.0 23.1 - 24.0 24.1 - 25.0 DROPULL I SIPËRM 25.1 - 26.0 > 26.0 Gini index (%) < 21.0 The boundaries of communes and municipalities have been designed for statistical purpose and 21.0 - 22.0 may not reflect exactly the territory of the local units. 22.1 - 23.0 23.1 - 24.0 24.1 - 25.0 Source: Population and Housing Gini coefficient Census is a measure 2011 of inequality The boundaries of communes and municipalities 25.1 - 26.0 have been designed for with values between 0 and 100. Living Standard Measurement Survey – LSMS 2012 > 26.0 the territory of the local unit statistical purpose and may not reflect exactly Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The prefecture counted 5,988 poor individuals and Odrie (11,839 ALL), which also showed in 2012. The biggest concentrations of the poor the lowest poverty rates in the prefecture, while were in the municipalities of Gjirokastër (1,350 the lowest level of consumption was in the poor individuals) and Përmet (418) and the municipalities of Krahës (8,751 ALL), Lopës commune of Qendër (363), and the lowest (8,761 ALL), and Qesarat (8,968 ALL). concentrations were in the communes of Odrie, The inequality measure in the prefecture was Pogon, and Zagori, where the number of poor 24.9 percent. The municipality of Gjirokastër in each was fewer than 20. showed the highest Gini coefficient, at 27.1 The average monthly per capita consumption percent, followed by the communes of Lazarat in the prefecture was 10,190 ALL in 2012, which (26.8 percent) and Dropull i Sipërm (25.9 was higher than the national average and higher percent), and the lowest coefficient of inequality than the average in the other prefectures. The was exhibited by the communes of Pogon (22.8 consumption level was highest in the communes percent), Sukë (23.5 percent) and Lopës (23.6 of Zagori (13,216 ALL), Pogon (11,936 ALL), percent). Portraits of Poverty and Inequality in Albania | Annexes 65 Korçë Prefecture The prefecture of Korçë is located in the central The poverty rate in the prefecture was 11.6 region. It is divided into 6 municipalities and 31 percent in 2012, which was below the national communes. The surface area of the prefecture average. The majority of communes had a is 3,711 square kilometers, and the population poverty rate between 8.0 percent and 16.0 is 219,649. The prefecture is bounded on the percent (map A7). The communes of Barmash, northeast and east by Republic of Macedonia, Liqenas, and Mollas and the municipality of on the south by Greece, on the southwest by Ersekë had rates at less than 8.0 percent. The the prefecture of Gjirokastër, on the west by the highest poverty rate in the prefecture belonged prefecture of Berat, and on the northwest by the to the commune of Çërravë (21.1 percent). prefecture of Elbasan. Map A7: Poverty Rate and Inequality, Korçë Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Prefecture of KORÇË PROPTISHT HUDENISHT VELÇAN TREBINJË POGRADEC BUÇIMAS ÇËRRAVË DARDHAS LIQENAS PIRG VRESHTAS GORË LIBONIK POJAN MALIQ PROGËR MOGLICË QENDËR QENDËR BILISHT VOSKOPOJË KORÇË BILISHT LEKAS HOÇISHT VOSKOP DRENOVË MOLLAJ VITHKUQ MIRAS MOLLAS ÇLIRIM QENDËR ERSEKË NOVOSELË Head Count ratio in (%). BARMASH < 8.0 8.0 - 12.0 12.1 - 16.0 LESKOVIK 16.1 - 20.0 LESKOVIK 20.1 - 24.0 24.1 - 28.0 > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 66 Portraits of Poverty and Inequality in Albania | Annexes Number of poor individuals b. Number of the poor Prefecture of KORÇË PROPTISHT HUDENISHT VELÇAN TREBINJË POGRADEC BUÇIMAS ÇËRRAVË DARDHAS LIQENAS PIRG VRESHTAS GORË LIBONIK POJAN MALIQ PROGËR MOGLICË QENDËR QENDËR BILISHT VOSKOPOJË BILISHT LEKAS KORÇË HOÇISHT VOSKOP DRENOVË MOLLAJ MIRAS VITHKUQ MOLLAS ÇLIRIM Persons QENDËR < 300 ERSEKË 300 - 600 NOVOSELË 601 - 900 901 - 1200 BARMASH 1201 - 1500 1501 - 1800 1801 - 2100 LESKOVIK 2101 - 3000 3001 - 9000 LESKOVIK > 9000 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and c. The average monthly per capita Living Standard consumption Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Per Capita Consumption Prefecture of KORÇË PROPTISHT HUDENISHT VELÇAN TREBINJË POGRADEC BUÇIMAS ÇËRRAVË DARDHAS LIQENAS PIRG VRESHTAS GORË LIBONIK POJAN MALIQ PROGËR MOGLICË QENDËR QENDËR BILISHT VOSKOPOJË BILISHT KORÇË LEKAS HOÇISHT VOSKOP DRENOVË MOLLAJ VITHKUQ MIRAS MOLLAS ÇLIRIM QENDËR ERSEKË NOVOSELË Per capita consumption (ALL/Month) < 7500 BARMASH 7500 - 8000 8001 - 8500 8501 - 9000 9001 - 9500 LESKOVIK 9501 - 10000 LESKOVIK 10001 - 10500 10501 - 11000 11001 - 11500 > 11500 Commune/Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Portraits of Poverty and Inequality in Albania | Annexes Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 67 Gini Coefficient d. Gini coefficient (inequality) Prefecture of KORÇË PROPTISHT HUDENISHT VELÇAN TREBINJË POGRADEC BUÇIMAS ÇËRRAVË DARDHAS LIQENAS PIRG VRESHTAS GORË LIBONIK POJAN MALIQ PROGËR MOGLICË QENDËR QENDËR BILISHT VOSKOPOJË BILISHT KORÇË LEKAS HOÇISHT VOSKOP DRENOVË MOLLAJ VITHKUQ MIRAS MOLLAS ÇLIRIM QENDËR ERSEKË NOVOSELË Gini index (%) BARMASH < 21.0 21.0 - 22.0 22.1 - 23.0 LESKOVIK 23.1 - 24.0 24.1 - 25.0 LESKOVIK 25.1 - 26.0 Gini coefficient is a measure of inequality with values between 0 and 100. > 26.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The largest concentrations of the poor in ALL) and Hudenisht (8,119 ALL), had the the prefecture were in Korçë and Pogradec lowest level of per capita consumption. municipalities and in Buçimas commune. The inequality level in the prefecture was The average monthly per capita 24.6 percent. The highest levels, over 26 consumption in the prefecture was 9,260 percent, were in the municipalities of Bilisht ALL. The communes of Liqenas, Mollas, and and Korçë and the commune of Voskopojë, Qendër (Kolonjë) and the municipalities of and the lowest level of inequality was Ersekë, Bilisht and Korçë showed the highest exhibited in the commune of Çlirim, at 22.6 levels of consumption, more than 10,000 percent. ALL. The communes of Çërravë (7,527 ALL), The prefecture of Kukës is located in the followed by the communes of Proptisht (8,068 mountain region. It is divided into 3 municipalities 68 Portraits of Poverty and Inequality in Albania | Annexes Kukës Prefecture and 24 communes. The surface area is 2,373 poverty rate in the country, around 22.0 percent, square kilometers, and the population is 85,250. in 2012. The communes of Grykë-Çajë, Kalis, The prefecture is bounded to the northeast and to Kolsh, Lekbibaj, Surroj, Topojan, Ujmisht, the east by Kosovo, to the south by the prefecture of and Zapod exhibited poverty rates above 28.0 Dibër, to the southwest by the prefecture of Lezhë, percent (map A8). The municipality of Bajram to the west by the prefecture of Shkodër, and to the Curri and the commune of Bytyç had the lowest northwest by Montenegro. poverty rates, respectively, 10.0 percent and 11.6 The prefecture of Kukës had the highest percent. Map A8: Poverty Rate and Inequality, Kukës Prefecture and communes, Poverty 2012 Head Count a. Poverty rate (headcount) Prefecture of KUKËS MARGEGAJ TROPOJË BAJRAM CURRI BUJAN LLUGAJ LEKBIBAJ BYTYÇ FIERZË GOLAJ KRUMË FAJZA GJINAJ TËRTHORE MALZI KUKËS KOLSH ZAPOD SHTIQËN SURROJ BICAJ SHISHTAVEC TOPOJAN Head Count ratio in (%). UJËMISHT < 8.0 ARRËN 8.0 - 12.0 GRYKË ÇAJË 12.1 - 16.0 BUSHTRICË 16.1 - 20.0 KALIS 20.1 - 24.0 24.1 - 28.0 > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 69 Prefecture of KUKËS MARGEGAJ TROPOJË Number of poor individuals b. Number of the poor Prefecture of KUKËS BUJAN BAJRAM CURRI LLUGAJ LEKBIBAJ TROPOJË MARGEGAJ BYTYÇ FIERZË GOLAJ BUJAN BAJRAM CURRI LLUGAJ KRUMË LEKBIBAJ FAJZA BYTYÇ GJINAJ FIERZË GOLAJ TËRTHORE MALZI KUKËS KOLSH KRUMË ZAPOD SHTIQËN FAJZA SURROJ GJINAJ BICAJ Persons SHISHTAVEC Persons TOPOJAN TËRTHORE MALZI KUKËS < 300 UJËMISHT KOLSH 300 - 600 ARRËN ZAPOD 601 - 900 SHTIQËN 901 - 1200 BUSHTRICË GRYKË ÇAJË 1201 - 1500 SURROJ 1501 - 1800 BICAJ KALIS Persons SHISHTAVEC 1801 - 2100 Persons TOPOJAN 2101 - 3000 < 300 UJËMISHT 300 - 600 3001 - 9000 ARRËN 601 - 900 901 - 1200 > 9000 BUSHTRICË GRYKË ÇAJË 1201 - 1500 1501 - 1800 Per Capita Consumption KALIS 1801 - 2100 Source: Population and Housing Census 2011 2101 - 3000 Prefecture ofdesigned The boundaries of communes and municipalities have been KUKËS Per for statisticalConsumption Capita purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. c. The average monthly per capita 3001 - 9000 consumption Prefecture of KUKËS > 9000 Source: MARGEGAJ TROPOJË The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. MARGEGAJ TROPOJË BAJRAM CURRI BUJAN LLUGAJ LEKBIBAJ BAJRAM CURRI BUJAN LLUGAJ BYTYÇ LEKBIBAJ FIERZË GOLAJ BYTYÇ FIERZË GOLAJ KRUMË FAJZA KRUMË GJINAJ TËRTHORE FAJZA GJINAJ MALZI KUKËS KOLSH TËRTHORE ZAPOD MALZI SHTIQËN KUKËS KOLSH ZAPOD SURROJ SHTIQËN Per capita consumption (ALL/Month) BICAJ SHISHTAVEC < 7500 TOPOJAN 7500 - 8000 SURROJ Per capita consumption (ALL/Month) BICAJ 8001 - 8500 UJËMISHT SHISHTAVEC < 7500 8501 - 9000 ARRËN TOPOJAN 7500 - 8000 9001 - 9500 8001 - 8500 GRYKË ÇAJË UJËMISHT 9501 - 10000 BUSHTRICË 8501 - 9000 ARRËN 10001 - 10500 9001 - 9500 10501 - 11000 KALIS GRYKË ÇAJË 9501 - 10000 BUSHTRICË 11001 - 11500 10001 - 10500 > 11500 10501 - 11000 KALIS Commune/Municipality Boundary 11001 - 11500 Source: > 11500 The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Commune/Municipality Boundary Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 70 Portraits of Poverty and Inequality in Albania | Annexes d. Gini coefficient (inequality) Gini Coefficient Prefecture of KUKËS MARGEGAJ TROPOJË BAJRAM CURRI BUJAN LLUGAJ LEKBIBAJ BYTYÇ FIERZË GOLAJ KRUMË FAJZA GJINAJ TËRTHORE MALZI KUKËS KOLSH ZAPOD SHTIQËN SURROJ Gini index (%) BICAJ SHISHTAVEC < 21.0 TOPOJAN 21.0 - 22.0 UJËMISHT 22.1 - 23.0 ARRËN 23.1 - 24.0 GRYKË ÇAJË 24.1 - 25.0 BUSHTRICË 25.1 - 26.0 > 26.0 KALIS Gini coefficient is a measure of inequality with values between 0 and 100. Source: The boundaries of communes and municipalities Population and Housing Census 2011 Source: Population and Housing Census 2011 Living Standard Measurement Survey – LSMS 2012 The boundaries of communes and municipalities have been designed for have been designed for statistical purpose and may not reflect exactly the territory of the local units. Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The largest concentration of the poor was in of Bajram Curri had the highest level of Kukës Municipality, the largest municipality consumption in the prefecture, above 8,700 in the area. ALL. Kukës was also the prefecture with the The level of inequality in the prefecture was lowest monthly per capita consumption below the national average, at 22.4 percent. in the country, 7,126 ALL. The majority of The lowest Gini coefficient was in Kalis the prefecture showed a consumption level Commune (19.7 percent), and the highest below 7,500 ALL. The commune of Kalis had value belonged to Fierzë Commune (25.0 the lowest level of consumption, 5,839 ALL. percent). The commune of Bytyç and the municipality Portraits of Poverty and Inequality in Albania | Annexes 71 Lezhë Prefecture The prefecture of Lezhë is located partly in of Dibër, on the south by the prefecture of the coastal region and partly in the central Durrës, and on the west by the Adriatic Sea. region. It is divided into 5 municipalities and The poverty rate in the prefecture was 17.2 16 communes. The surface area is 1,588 square percent in 2012. The communes of Balldren i Ri, kilometers, and the population is 133,275. The Fan, and Ungrej had the highest poverty rates in prefecture is bounded on the north by the the prefecture, at more than 20.0 percent (map prefecture of Shkodër, on the northwest by the A9). The municipality of Rrëshen had the lowest prefecture of Kukës, on the east by the prefecture poverty rate, at 11.2 percent. Map A9: Poverty Rate and Inequality, Lezhë Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Poverty Head Count Prefecture of LEZHË Prefecture of LEZHË FAN FAN DAJÇ DAJÇ BLINISHT BLINISHT UNGREJ KAÇINAR UNGREJ KAÇINAR BALLDREN I RI BALLDREN I RI OROSH OROSH KALLMET KALLMET KOLÇ KOLÇ LEZHË RRËSHEN RRËSHEN SELITË LEZHË RUBIK SELITË RUBIK SHËNGJIN SHËNGJIN KTHJELLË KTHJELLË ZEJMEN ZEJMEN Head Count ratio in (%). Head Count ratio in (%). SHËNKOLL SHËNKOLL < 8.0 < 8.0 8.0 - 12.0 8.0 - 12.0 MILOT MILOT LAÇ 12.1 - 16.0 12.1 - 16.0 LAÇ FUSHË KUQE FUSHË KUQE 16.1 - 20.0 16.1 - 20.0 20.1 - 24.0 20.1 - 24.0 MAMURRAS MAMURRAS 24.1 - 28.0 24.1 - 28.0 > 28.0 > 28.0 Source: The boundaries of communes and municipalities Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 72 Portraits of Poverty and Inequality in Albania | Annexes Number of poor individuals Number of poor individuals b. Number of the poor Prefecture of LEZHË Prefecture of LEZHË FAN FAN DAJÇ DAJÇ BLINISHT BLINISHT KAÇINAR KAÇINAR UNGREJ UNGREJ BALLDREN I RI BALLDREN I RI OROSH OROSH KALLMET KALLMET SHËNGJIN SHËNGJIN KOLÇ KOLÇ LEZHË LEZHË RRËSHEN SELITË SELITË RUBIK RUBIK RRËSHEN Persons Persons KTHJELLË KTHJELLË ZEJMEN ZEJMEN < 300 < 300 300 - 600 300 - 600 SHËNKOLL SHËNKOLL 601 - 900 601 - 900 901 - 1200 901 - 1200 1201 - 1500 1201 - 1500 1501 - 1800 1501 - 1800 MILOT MILOT LAÇ LAÇ 1801 - 2100 1801 - 2100 FUSHË KUQE FUSHË KUQE 2101 - 3000 2101 - 3000 3001 - 9000 3001 - 9000 MAMURRAS MAMURRAS > 9000 > 9000 Source: Source: The boundaries of communesThe municipalities andboundaries of communes and municipalities Population and Housing Census 2011 and Housing Census 2011 Population havepurpose have been designed for statistical and for statistical purpose and been designed Survey Living Standard MeasurementLiving 2012 – LSMSMeasurement Standard Survey – LSMS 2012 may not reflect exactly the territory may notof the local reflect units. the territory of the local units. exactly Per Capita Consumption c. The average monthly per capita consumption Per Capita Consumption Prefecture of LEZHË Prefecture of LEZHË FAN DAJÇ FAN BLINISHT UNGREJ KAÇINAR DAJÇ BALLDREN I RI BLINISHT OROSH UNGREJ KAÇINAR KALLMET BALLDREN I RI OROSH KALLMET LEZHË KOLÇ RRËSHEN SELITË RUBIK SHËNGJIN LEZHË KOLÇ RRËSHEN SELITË RUBIK KTHJELLË Per capita consumption (ALL/Month) ZEJMEN SHËNGJIN < 7500 SHËNKOLL KTHJELLË Per capita consumption (ALL/Month) 7500 - 8000 ZEJMEN < 7500 8001 - 8500 SHËNKOLL 7500 - 8000 8501 - 9000 9001 - 9500 MILOT 8001 - 8500 LAÇ 8501 - 9000 9501 - 10000 FUSHË KUQE 10001 - 10500 9001 - 9500 MILOT 10501 - 11000 LAÇ 9501 - 10000 FUSHË KUQE 11001 - 11500 MAMURRAS 10001 - 10500 > 11500 10501 - 11000 11001 - 11500 Commune/Municipality Boundary MAMURRAS > 11500 Source: The boundaries of communes and municipalities Population and Housing Census 2011 Commune/Municipality Boundary have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 73 Gini Coefficient Gini Coefficient Prefecture of LEZHË d. Gini coefficient (inequality) Prefecture of LEZHË FAN FAN DAJÇ BLINISHT UNGREJ KAÇINAR DAJÇ BLINISHT BALLDREN I RI UNGREJ KAÇINAR OROSH BALLDREN I RI KALLMET OROSH KALLMET KOLÇ LEZHË RRËSHEN SELITË RUBIK KOLÇ SHËNGJIN RRËSHEN LEZHË SELITË RUBIK SHËNGJIN KTHJELLË ZEJMEN KTHJELLË ZEJMEN SHËNKOLL SHËNKOLL Gini index (%) < 21.0 Gini index (%) MILOT LAÇ 21.0 - 22.0 < 21.0 FUSHË KUQE MILOT 22.1 - 23.0 LAÇ 21.0 - 22.0 23.1 - 24.0 FUSHË KUQE 22.1 - 23.0 24.1 - 25.0 MAMURRAS Gini coefficient is a measure of inequality 23.1 - 24.0 with values between 0 and 100. 25.1 - 26.0 24.1 - 25.0 Gini coefficient is a measure of inequality MAMURRAS > 26.0 with values between 0 and 100. 25.1 - 26.0 > 26.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Source: Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local The boundaries units. of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Source: Population and Housing Census 2011 Living Standard Measurement Survey – LSMS 2012 The boundaries of communes and municipalities have been designed for may not reflect exactly the territory of the local units. Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The number of the poor in the prefecture was capita consumption. The municipality of Lezhë 22,926 in 2012. The municipalities of Laç and had the highest level of consumption (9,812 Mamurras and the commune of Shënkoll had ALL), followed by the municipality of Rrëshen the largest concentrations of poor within the (9,180 ALL) and the commune of Shëngjin prefecture. Kaçinar, Orosh, and Selitë are the (9,128 ALL). communes with the smallest concentrations of The level of inequality in the prefecture was the poor. 24.6 percent. The municipality of Lezhë and the The average monthly per capita consumption commune of Shëngjin had levels of inequality in the prefecture was 8,142 ALL. The communes above 26 percent. The lowest level of inequality of Fan and Ungrej had the lowest average per was in the commune of Ungrej, at 22.1 percent. 74 Portraits of Poverty and Inequality in Albania | Annexes Shkodër Prefecture The prefecture of Shkodër is located in the The poverty rate in the prefecture was 14.0 northwest of Albania in the central region. percent in 2012. The highest poverty rate in It is divided into 5 municipalities and 28 the prefecture was recorded in the commune communes. The surface area of the prefecture of Blerim (23.7 percent), followed by the is 1,588 square kilometers, with a population of communes of Qelëz (22.1 percent), Fierzë (22.1 213,706 inhabitants. The prefecture is bounded percent), and Qafë-Mali (21.2 percent) (map on the north by Montenegro, on the east by A10). The lowest poverty rate in the prefecture the prefecture of Kukës, on the south by the was in the commune of Shalë (8.7 percent), prefecture of Lezhë, and on the west by the followed by the commune of Dajç (9.1 percent) Adriatic Sea. and the municipality of Koplik (9.3 percent). Poverty Head Count Prefecture Map A10: SHKODËR Rate and Inequality, Shkodër Prefecture and communes, 2012 ofPoverty a. Poverty rate (headcount) KELMEND Poverty Head Count Prefecture of SHKODËR SHALË L KELMEND PULT SHOSH FIERZË IBALLË TEMAL BLERIM SHALË POSTRIBË KASTRAT SHLLAK SHKREL QELËZ FUSHË ARRËZ QAFË MALI PUKË RRAPË VAU I DEJËS PULT QENDËR QERRET KOPLIK QENDËR SHOSH FIERZË GRUEMIRË VIG-MNELË AJMEL IBALLË GJEGJAN TEMAL BLERIM POSTRIBË RRETHINAT SHLLAK QELËZ FUSHË ARRËZ SHKODËR GURI I ZI QAFË MALI Head Count ratio in (%). RRETHINAT ANA E MALIT PUKË RRAPË VAU I DEJËS < 8.0 BËRDICË QERRET 8.0 - 12.0 12.1 - 16.0 DAJÇ VIG-MNELË BUSHAT HAJMEL 16.1 - 20.0 GJEGJAN 20.1 - 24.0 VELIPOJË 24.1 - 28.0 > 28.0 The boundaries of communes and municipalities have been designed for statistical purpose and Head Count ratio in (%). may not reflect exactly the territory of the local units. < 8.0 8.0 - 12.0 12.1 - 16.0 16.1 - 20.0 20.1 - 24.0 24.1 - 28.0 Portraits of Poverty and Inequality in Albania | Annexes > 28.0 75 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Prefecture of SHKODËR ELMEND b. Number of the poor SHALË Number of poor individuals Prefecture of SHKODËR KELMEND PULT SHOSH FIERZË IBALLË TEMAL BË BLERIM SHLLAK SHALË QELËZ KASTRAT FUSHË ARRËZ SHKREL QAFË MALI RRAPË VAU I DEJËS PUKË QERRET PULT QENDËR KOPLIK QENDËR SHOSH FIERZË VIG-MNELË GJEGJAN GRUEMIRË IBALLË TEMAL POSTRIBË BLERIM RRETHINAT Persons SHLLAK QELËZ FUSHË ARRËZ < 300 SHKODËR 300 - 600 RRETHINAT 601 - 900 GURI I ZI RRAPË QAFË MALI ANA E MALIT VAU I DEJËS PUKË 901 - 1200 1201 - 1500 BËRDICË QERRET 1501 - 1800 DAJÇ 1801 - 2100 BUSHAT VIG-MNELË HAJMEL GJEGJAN Per Capita Consumption 2101 - 3000 3001 - 9000 VELIPOJË Prefecture of SHKODËR > 9000 Persons The boundaries of communes and municipalities < 300 have been designed for statistical purpose and 300 - 600 may not reflect exactly the territory of the local units. 601 - 900 END 901 - 1200 c. The average monthly per capita consumption 1201 - 1500 1501 - 1800 Per Capita Consumption 1801 - 2100 2101 - 3000 Prefecture of SHKODËR 3001 - 9000 SHALË > 9000 Source: The boundaries of communes and municipalities KELMEND Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. LT SHOSH FIERZË IBALLË TEMAL BLERIM SHALË KASTRAT SHLLAK SHKREL QELËZ FUSHË ARRËZ QAFË MALI PUKË RRAPË PULT U I DEJËS QENDËR KOPLIK QENDËR SHOSH FIERZË QERRET GRUEMIRË IG-MNELË IBALLË GJEGJAN TEMAL BLERIM POSTRIBË RRETHINAT SHLLAK QELËZ Per capita consumption (ALL/Month) FUSHË ARRËZ < 7500 SHKODËR 7500 - 8000 GURI I ZI QAFË MALI RRETHINAT RRAPË ANA E MALIT PUKË 8001 - 8500 VAU I DEJËS 8501 - 9000 BËRDICË QERRET 9001 - 9500 DAJÇ 9501 - 10000 BUSHAT HAJMEL VIG-MNELË GJEGJAN 10001 - 10500 10501 - 11000 VELIPOJË 11001 - 11500 Per capita consumption (ALL/Month) > 11500 < 7500 Commune/Municipality Boundary 7500 - 8000 The boundaries of communes and municipalities 8001 - 8500 have been designed for statistical purpose and may not reflect exactly the territory of the local units. 8501 - 9000 76 Portraits of Poverty and Inequality in Albania 9001 - 9500 | Annexes 9501 - 10000 10001 - 10500 10501 - 11000 11001 - 11500 > 11500 Gini Coefficient Prefecture of SHKODËR KELMEND d. Gini coefficient (inequality) Gini Coefficient Prefecture of SHKODËR SHALË SHKREL KELMEND PULT SHOSH FIERZË IBALLË TEMAL SHALË BLERIM POSTRIBË KASTRAT SHKREL SHLLAK QELËZ FUSHË ARRËZ I I ZI QAFË MALI PUKË RRAPË PULT VAU I DEJËS QENDËR KOPLIK QENDËR SHOSH FIERZË QERRET GRUEMIRË IBALLË VIG-MNELË HAJMEL TEMAL GJEGJAN BLERIM POSTRIBË RRETHINAT SHLLAK QELËZ FUSHË ARRËZ SHKODËR GURI I ZI QAFË MALI Gini index (%) ANA E MALIT RRETHINAT PUKË RRAPË VAU I DEJËS < 21.0 BËRDICË QERRET 21.0 - 22.0 22.1 - 23.0 DAJÇ BUSHAT VIG-MNELË HAJMEL 23.1 - 24.0 GJEGJAN 24.1 - 25.0 25.1 - 26.0 VELIPOJË > 26.0 The boundaries of communes and municipalities Gini index (%) have been designed for statistical purpose and S 2012 may not reflect exactly the territory of the local units. < 21.0 21.0 - 22.0 22.1 - 23.0 23.1 - 24.0 24.1 - 25.0 Gini coefficient is a measure of inequality with values between 0 and 100. 25.1 - 26.0 > 26.0 Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey Source: – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. The prefecture counted 29,912 poor individuals 10,000 ALL. The lowest level of consumption in 2012. The municipality of Shkodër and the was in the communes of Blerim, Fierzë, and commune of Rrethinat had the highest number Qelëz, less than 7,500 ALL. of poor in the prefecture, respectively, 9,199 The level of inequality in the prefecture was poor and 3,590 poor. 25.1 percent. The highest level of inequality, over The average monthly per capita consumption 26 percent, was in the communes of Bushat, in the prefecture was 8,664 ALL. The highest Dajç, Kastrat, Velipojë and in the municipalities level of consumption was in the communes of of Koplik and Shkodër. The lowest level of Dajç and Shalë and the municipality of Koplik, inequality was in the commune of Qelëz (22.6 where the level of consumption averaged over percent). Portraits of Poverty and Inequality in Albania | Annexes 77 Tirana Prefecture The prefecture of Tirana is located in the remains attractive because of the opportunities central, coastal, and Tirana regions. It is divided in employment, education, and so on. However, into 5 municipalities and 24 communes. The migrants often face difficulty in becoming surface area of the prefecture is 1,586 square assimilated (INSTAT and IOM 2014). kilometers, and the population is 742,515, The poverty rate in the prefecture was 12.7 around one-fifth of the population of Albania. percent in 2012. The highest poverty rate was The prefecture is bounded in the north by in the municipality of Kamëz (25.2 percent) the prefecture of Durrës, in the northeast and the commune of Krrabë (24.9 percent) by the prefecture of Dibër, in the southeast (map A11). The lowest poverty rate was in by the prefecture of Elbasan, in the south by the municipalities of Farkë and Tirana, at less the prefecture of Fier, and in the west by the than 10.0 percent. Adriatic Sea. Tirana, the capital of Albania, Map A11: Poverty Rate and Inequality, Tirana Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Prefecture of TIRANË PREZË ZALL BASTAR ZALL HERR BËRXULLË KAMËZ VORË PASKUQAN KASHAR SHËNGJERGJ TIRANË DAJT FARKË VAQARR NDROQ PETRELË BËRZHITË GOLEM PEZË KËRRABË HELMËS BALDUSHK KAVAJË SYNEJ Head Count ratio (%). LUZ I VOGEL < 8.0 LEKAJ 8.0 - 12.0 SINABALLAJ KRYEVIDH 12.1 - 16.0 16.1 - 20.0 RROGOZHINË GOSË 20.1 - 24.0 24.1 - 28.0 > 28.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 78 Portraits of Poverty and Inequality in Albania | Annexes b. Number of the poor Number of poor individuals Prefecture of TIRANË ZALL BASTAR PREZË ZALL HERR VORË KAMËZ BËRXULLË PASKUQAN KASHAR DAJT TIRANË SHËNGJERGJ FARKË VAQARR NDROQ PETRELË BËRZHITË GOLEM PEZË KËRRABË Persons KAVAJË HELMËS BALDUSHK < 300 300 - 600 SYNEJ 601 - 900 901 - 1200 LUZ I VOGEL LEKAJ 1201 - 1500 1501 - 1800 SINABALLAJ KRYEVIDH 1801 - 2100 RROGOZHINË 2101 - 3000 GOSË 3001 - 9000 > 9000 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. c. The average monthly per capita consumption Per Capita Consumption Prefecture of TIRANË PREZË ZALL BASTAR ZALL HERR BËRXULLË KAMËZ VORË PASKUQAN KASHAR SHËNGJERGJ TIRANË DAJT FARKË VAQARR NDROQ PETRELË BËRZHITË GOLEM PEZË KËRRABË Per capita consumption (ALL/Month) HELMËS BALDUSHK KAVAJË < 7500 7500 - 8000 SYNEJ 8001 - 8500 LUZ I VOGEL LEKAJ 8501 - 9000 9001 - 9500 SINABALLAJ KRYEVIDH 9501 - 10000 10001 - 10500 RROGOZHINË GOSË 10501 - 11000 11001 - 11500 > 11500 Commune/Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Portraits of Poverty and Inequality in Albania | Annexes 79 d. Gini coefficient (inequality) Gini Coefficient Prefecture of TIRANË PREZË ZALL BASTAR ZALL HERR BËRXULLË KAMËZ VORË PASKUQAN KASHAR SHËNGJERGJ TIRANË DAJT FARKË VAQARR NDROQ PETRELË BËRZHITË GOLEM PEZË KËRRABË HELMËS BALDUSHK KAVAJË SYNEJ LUZ I VOGEL Gini index (%) LEKAJ < 21.0 SINABALLAJ KRYEVIDH 21.0 - 22.0 22.1 - 23.0 RROGOZHINË GOSË 23.1 - 24.0 Gini coefficient is a measure of inequality 24.1 - 25.0 with values between 0 and 100. 25.1 - 26.0 > 26.0 Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The number of the poor in the prefecture was consumption were in the municipalities of 83,845 in 2012. The municipalities of Kamëz Farkë (11,001 ALL) and Tirana (10,647 ALL). and Tirana and the commune of Paskuqan The lowest consumption levels were in the had the largest concentrations of poor. An area communes of Kamëz, Krrabë and Sinaballaj, at that is less poor in terms of the poverty rate less than 7,500 ALL. may be poorer in terms of the number of the The level of inequality in the prefecture was poor. This was the case, for example, of Tirana 25.0 percent. The areas with the highest level of Municipality. inequality in the prefecture, at over 26 percent, were The average monthly per capita consumption Dajt, Farkë, Kashar, Kavajë, Prezë and Tirana. in the prefecture of Tirana was lek 8,584 in 2012. The highest levels of monthly per capita 80 Portraits of Poverty and Inequality in Albania | Annexes Minimunicipalities of Tirana The municipality of Tirana is divided into (Tirana 4), Kombinati (6), and Laprakë (11) 11 neighborhoods or minimunicipalities. The (map A12). The lowest poverty rate, below poverty rate in the municipality was 9.2 percent 8.0 percent, was in Qyteti Studenti (Tirana 2), in 2012. The highest poverty rates, at more than Tirana e Re (5), and City Center (10). 12.0 percent, were in Allias and Kinostudio Map A12: Poverty Rate and Inequality, minimunicipalities of Tirana, 2012 a. Poverty rate (headcount) Poverty Head Count Poverty Head Count Municipality of TIRANË Municipality of TIRANË 4 4 11 11 8 8 9 9 3 3 7 7 1 10 1 10 5 5 Head count ratio (%)count ratio (%) Head 2 2 6 6 < 8.0 < 8.0 8.0 - 10.0 8.0 - 10.0 10.1 - 12.0 10.1 - 12.0 > 12.0 > 12.0 Mini municipality Boundary Mini municipality Boundary Source: Source: communes The boundaries ofThe andof municipalities Population and Housing Census boundaries communes and municipalities Population and 2011 Housing Census 2011 have been designed forbeen statistical have purpose designed and for statistical purpose and Living Standard Measurement Survey Living Standard – LSMS 2012 Measurement Survey – LSMS 2012 may not reflect exactly the reflect may not territory of thethe exactly local units. of the local units. territory Portraits of Poverty and Inequality in Albania | Annexes 81 b. Number of the poor Number of poor individuals Number of poor individuals Municipality of TIRANË Municipality of TIRANË 4 4 11 11 8 9 8 9 3 3 7 7 1 10 1 10 5 2 6 5 2 6 Persons Persons < 2000 < 2000 2000 - 3000 2000 - 3000 3001 - 5000 3001 - 5000 > 5000 > 5000 Mini_Municipality Boundary Mini_Municipality Boundary Source: The boundaries of communes and municipalities Source: Census 2011 Population and Housing have been designed for statistical The boundaries ofpurpose communes andand municipalities Measurement Living Standard Population andSurvey – LSMS Housing 2012 Census 2011 may not reflect exactly the territory have been designed offor the local units. statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. c. The average monthly per capita consumption Per capita consumption Per capita consumption 4 11 4 11 8 9 8 9 3 3 7 1 710 1 10 5 5 Per capita consumption ( ALL/month ) 2 Per capita consumption ( ALL/month ) 6 2 < 9000 6 < 9000 9000 - 10000 9000 - 10000 10001 - 11500 10001 - 11500 > 11500 > 11500 Mini - Municipality Boundary Mini - Municipality Boundary Source: The boundaries of communes and municipalities Source: 2011 Population and Housing Census have been designed for statistical purpose Living Standard Measurement Survey – LSMS 2012 The boundaries ofand communes and municipalities Population and Housing Census 2011 may not reflect exactly the territory have beenof the localfor designed units. statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 82 Portraits of Poverty and Inequality in Albania | Annexes d. Gini coefficient (inequality) Gini Coefficient Gini Coefficient Municipality of TIRANË Municipality of TIRANË 4 11 8 4 11 9 8 3 9 3 7 1 10 7 1 10 5 5 2 Gini Index (%) 6 2 < 26.5 Gini Index (%) 6 26.5 - 27.0 < 26.5 27.1 - 27.2 26.5 - 27.0 > 27.2 Gini coefficient is a measure of inequality 27.1 - 27.2 with values between 0 and 100. Mini municipality Boundary > 27.2 Gini coefficient is a measure of inequality Source: with values between 0 and 100. Mini municipality Boundary The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The relatively smallest number of poor were 10). The lowest level of consumption was in living in the City Center (Tirana 10). Kombinati (Tirana 6). The level of inequality in The average monthly per capita consumption the municipality of Tirana was 27.7 percent.The in the municipality of Tirana was 10,647 ALL in highest level of inequality, over 27 percent, was 2012. The highest levels of monthly per capita in 21-Dhjetori (Tirana 7), Medreseja and Selvia consumption, more than 11,500 ALL, were in (Tirana 8). The lowest level of inequality was in Tirana e Re (Tirana 5) and City Center (Tirana Kombinati (Tirana 6). Portraits of Poverty and Inequality in Albania | Annexes 83 Vlorë Prefecture The prefecture of Vlorë is located in the coastal the Ionian Sea and the Adriatic Sea. The poverty region. It is divided into 7 municipalities and 19 rate in the prefecture was 13.2 percent in 2012. communes. The surface area of the prefecture The commune of Vllahinë is the poorest area is 2,706 square kilometers, and the population in the prefecture, with a poverty rate of 23.0 is 174,524. The prefecture is bounded on the percent (map A13). The lowest poverty rates north by the prefecture of Fier, on the east by were in the communes of Dhivër and Livadhja, the prefecture of Gjirokastër, on the south by at less than 7.0 percent. Greece, and on the southwest and the west by Map A13: Poverty Rate and Inequality, Vlorë Prefecture and communes, 2012 a. Poverty rate (headcount) Poverty Head Count Prefecture of VLORË NOVOSELË SELENICË QENDËR SHUSHICË ARMEN QENDËR VLORË VLLAHINË SEVASTER KOTE ORIKUM BRATAJ VRANISHT HIMARË VERGO LUKOVË DELVINË MESOPOTAM FINIQ SARANDË Head Count ratio in (%). DHIVËR < 8.0 ALIKO 8.0 - 12.0 12.1 - 16.0 KSAMIL LIVADHJA 16.1 - 20.0 20.1 - 24.0 XARRË MARKAT 24.1 - 28.0 > 28.0 KONISPOL Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 84 Portraits of Poverty and Inequality in Albania | Annexes Number of poor individuals b. Number of the poor NOVOSELË Prefecture of VLORË SHUSHICË SELENICË QENDËR ARMEN QENDËR VLORË VLLAHINË SEVASTER KOTE ORIKUM BRATAJ VRANISHT HIMARË LUKOVË VERGO DELVINË Persons MESOPOTAM SARANDË FINIQ < 300 300 - 600 601 - 900 DHIVËR ALIKO 901 - 1200 1201 - 1500 KSAMIL 1501 - 1800 LIVADHJA 1801 - 2100 XARRË 2101 - 3000 MARKAT 3001 - 9000 KONISPOL > 9000 c. The average monthly per capita Source: consumption Per Capita Consumption The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Prefecture of VLORË NOVOSELË SELENICË QENDËR SHUSHICË ARMEN QENDËR VLORË VLLAHINË SEVASTER KOTE ORIKUM BRATAJ VRANISHT HIMARË VERGO LUKOVË DELVINË Per capita consumption (ALL/Month) MESOPOTAM FINIQ < 7500 SARANDË 7500 - 8000 8001 - 8500 DHIVËR ALIKO 8501 - 9000 9001 - 9500 KSAMIL LIVADHJA 9501 - 10000 10001 - 10500 10501 - 11000 XARRË MARKAT 11001 - 11500 > 11500 KONISPOL Commune/Municipality Boundary Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Portraits of Poverty and Inequality in Albania | Annexes Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. 85 d. Gini coefficient (inequality) Gini Coefficient Prefecture of VLORË NOVOSELË SELENICË SHUSHICË QENDËR ARMEN QENDËR VLORË VLLAHINË SEVASTER KOTE ORIKUM BRATAJ VRANISHT HIMARË VERGO LUKOVË DELVINË Gini index (%) MESOPOTAM FINIQ < 21.0 SARANDË 21.0 - 22.0 DHIVËR 22.1 - 23.0 ALIKO 23.1 - 24.0 24.1 - 25.0 KSAMIL LIVADHJA 25.1 - 26.0 XARRË > 26.0 MARKAT KONISPOL Gini coefficient is a measure of inequality with values between 0 and 100. Source: The boundaries of communes and municipalities Population and Housing Census 2011 have been designed for statistical purpose and Living Standard Measurement Survey – LSMS 2012 may not reflect exactly the territory of the local units. Source: Population and Housing Census 2011 The boundaries of communes and municipalities have been designed for Living Standard Measurement Survey – LSMS 2012 statistical purpose and may not reflect exactly the territory of the local unit The prefecture counted 23,092 poor individuals (10,134 ALL). The lowest level of consumption was in 2012. The municipality of Vlorë had the largest in the commune of Vllahinë, less than 7,500 ALL. number of the poor in the prefecture, at 9,699. The level of inequality in the prefecture was The average monthly per capita consumption in 25.1 percent. The highest level of inequality, over the prefecture of Vlorë was 8,867 ALL. The highest 27 percent, was in the municipalities of Himarë, levels of monthly per capita consumption were found Sarandë, and Vlorë. The lowest level of inequality in the communes of Livadhja (10,843 ALL) and was in the communes of Sevaster and Xarrë, at less Dhivër (10,502 ALL) and the municipality of Sarandë than 23 percent. 86 Portraits of Poverty and Inequality in Albania | Annexes Annex A: Updates to the First Albania Poverty Maps, 2001–2010 The analyses of poverty and inequality maps during the initial mapping exercise in Albania was carried out by Betti, Neri, and Ballini (2003), based on the methodology fully described in Elbers, Lanjouw, and Lanjouw (2003). This methodology (the ELL) combined data of the 2001 census and the 2002 LSMS to produce detailed disaggregated maps that described the spatial distribution of poverty and inequality in the country. Dabalen and Ferrè (2008) updated the first poverty mapping by means of the 2001 census and the new round of the LSMS, the Albania 2005 LSMS. The contribution of this work relied on a reweighting scheme. Specifically, the method, which followed the research of Lemieux (2002), involved the construction of a counterfactual consumption distribution for the old household survey, in our case the Albania LSMS 2002, using the information contained in the latest survey, the Albania LSMS 2005. The method projected how the consumption distribution for the 2002 LSMS (which was conducted about the time the 2001 census was done) would look if the parameters (the coefficients of a consumption model) and the distribution of the characteristics of the sample and, therefore, population were as reported in the 2005 survey. The derived counterfactual distribution, together with the 2001 census, were then used to obtain an updated poverty map of the country using the ELL methodology as applied in the original poverty and inequality mapping by Betti, Neri, and Ballini (2003). Later, Betti et al. (2010) updated the poverty mapping in Albania by using the procedure proposed by Dabalen and Ferrè (2008). They constructed a counterfactual distribution of the monetary variable for the LSMS 2002 using information from the latest survey, the Albania LSMS 2008, and the 2001 census. They then applied the original ELL method on the basis of the counterfactual distribution, which led to a monetary distribution in the 2001 census data related to the 2008 LSMS. Portraits of Poverty and Inequality in Albania | Annexes 87 Annex B: The ELL Methodology: Innovations and Criticisms ELL has been used to produce more than 100 poverty maps worldwide. Its wide popularity partly derives from the development of the PovMap software by the World Bank that has made the implementation of the technique more tractable.7 Box 1: The Small Area Estimation Method Developed by ELL (2003) The method proposed by ELL has three stages. First, a set of variables deemed to have similar distributions in the survey and the census are identified. n yc Second, a model of log per capita consumption expenditure ( l h ) is estimated in the survey data based on the identified variables: ′ ′ h = Xc n yc l h b + Z g + uc h (1) ′ where X c h is the vector of explanatory variables for household h in cluster c, β is the vector ′ of regression coefficients, Z is the vector of location specific variables, g is the vector of coefficients, and u c h is the error term due to the discrepancy between the predicted ′ household consumption and the actual value. X c h Is household level variables that have ′ similar distributions in both survey and census, while Z include location specific averages of variables found in census, and potentially other external variables available at local levels for the entire country, as for instance GIS based variables. This error term is decomposed into two independent components: u c h = h c + ec h where h c is cluster-specific effect and e c h a household-specific effect. This error structure allows for both a location effect – common to all households in the same area—and heteroskedasticity in the household-specific errors. The location effect can be any level. In the third part of the analysis, poverty estimates and their standard errors are computed. There are two sources of errors involved in the estimation process: errors in the estimated regression coefficients ( bˆ,g ˆ ) and the disturbance terms, both of which affect poverty estimates and the level of their accuracy. ELL propose a way to properly calculate poverty estimates as well as their standard errors while taking into account these sources of bias. A simulated value of expenditure ′ˆ ′ˆ for each census household is calculated with predicted log expenditure X c h b + Z g and random draws from the estimated distributions of the disturbance terms, h c and e c h . In the case of Albania, these simulations are repeated 200 times. For any given location (such as a district or a commune), the mean across the 200 simulations of a poverty statistic provides a point estimate of the statistic, and the standard deviation provides an estimate of the standard error. A so-called Alpha model (see Elbers et al., 2003) is also used in these models. 7. PovMap is a freeware available at Software for Poverty Mapping (database), World Bank, Washington, DC, http://go.worldbank. org/QG9L6V7P20. In addition to the software, the site also includes links to the large literature on the topic. This includes examples of applications around the globe. See also Zhao and Lanjouw (2009). 88 Portraits of Poverty and Inequality in Albania | Annexes Box 1 recaps the methodology and the stepwise approach. There have been several recent methodological developments within the literature on small area estimation that have been included in the new version of the PovMap software, PovMap2. The methodological improvements include estimations via the empirical best method, an estimation method proposed in Molina and Rao (2010) that utilizes existing information in the household survey more efficiently and has a particular advantage if surveys cover a large number of PSUs. Further improvements also include the options of utilizing the empirical best method based on an approximated empirical distribution (the normal mixtures approach) instead of an assumed distribution (Elbers and van der Weide 2014). These recent methodological contributions are used in the Albanian poverty map. A major advantage of the ELL-based approach is that, in addition to estimating key indicators such as poverty rates, the depth of poverty, and a range of inequality measures, it also estimates the standard errors associated with these indicators. The importance of standard errors becomes a critical aspect in gauging how much trust to have in the predictions. A prediction that, say, 15 percent of an area is poor is not so valuable if it comes with a 95 percent confidence interval of, say, 0 percent to 30 percent. As illustration of standard errors that are more useful, the 95 percent confidence interval for the national poverty estimate of the LSMS survey is 12.5 percent to 16.1 percent poor. A number of factors influence the standard errors of our poverty estimates. This includes sampling and measurement errors in the survey, which are beyond the control of the process for the construction of the poverty maps. Other aspects that we do have some control over are the precision of the consumption model, the spatial level at which cluster effects are estimated, and the number of households in each spatial area for which we estimate poverty. In this application 200 iterations were made to simulate standard errors. Tarozzi and Deaton (2008) highlight some concerns with the ELL methodology. Notably, they show that, under certain circumstances, the ELL method can result in an overly optimistic assessment of the standard errors of the local poverty estimates. The specific concerns raised by Tarozzi and Deaton (2008) can be summarized as follows. First, differences in consumption patterns not captured in the model can bias the poverty estimates and the standard errors. The ELL method estimates a consumption model that is assumed to apply to all households within each model. The implicit assumption is that the relationship between household expenditures and their correlates are the same for all households and that all remaining differences arise because of nonstructural factors. This is not a minor assumption and is explicitly acknowledged as such in Elbers, Lanjouw, and Lanjouw (2003). However, Elbers, Lanjouw, and Lanjouw (2003) provide evidence that the concern does not have large practical implications. Second, Tarozzi and Deaton (2008) caution that the misspecification in the error structure can lead to underestimation of standard errors. They show that, under some conditions, ignoring the spatial correlation can cause a bias in the standard errors of the poverty estimates. Portraits of Poverty and Inequality in Albania | Annexes 89 Annex D: Differences and Similarities: 2002 and 2012 Maps This appendix makes a comparison of the poverty maps of 2002 and the new poverty maps. The comparison has two purposes: (1) to document the technical differences and similarities in the estimations and (2) to provide insights on the degree of comparability between the two sets of maps. The estimations in both years rely on the same methodology, with several variations in implementation, including some that may impact the results. Nonetheless, the two sets of maps should be reflect reasonably unbiased estimates of poverty at two points in time, which should be comparable. However, we do not have any evidence or experience on how much the variations may impact the estimated results. For this reason, any analysis based on the estimates should be careful and systematic, and any comparison of individual municipalities is not recommended. This is particularly true in the case of the communes because many communes have small populations, and the estimates are therefore associated with a larger degree of uncertainty. Data Both mapping exercises were able to rely on a survey and a census close in time: the 2002 mapping process depended on the 2001 census, while the 2012 mapping process depended on the 2011 census. Estimation methodology and implementation Both sets of maps rely on the ELL estimation method. A standard implementation of the ELL includes the following stages: Stage 0: Comparability of survey and census Stage 1: Regression modeling Stage 2: Simulation and estimation Below, the maps are compared at each stage. Stage 0: Comparability of survey and census During stage 0, the basic explanatory variables are identified through a detailed comparison of the questionnaires and the distributions of the variables in each data source. Normally, only variables found to be similar in both data sources are utilized in the regression models. In 2002, only a few variables were found to have similar distributions in the survey and the census, and stage 0 was not undertaken. Variations in the distributions could have two causes: Changes in the economy between the census and the survey 90 Portraits of Poverty and Inequality in Albania | Annexes Differences in the survey instruments or in the survey implementation If the variations arose because of changes in the economy, this would mean that the estimated levels of poverty reflect the situation in the year of the census (2001) instead of the year of the survey (2002). With just one year between the two data sources, it seems unlikely that most variables underwent massive changes. If the variations arose because of differences in the instruments or the implementation, this means the estimates may be biased. Imagine, for instance, that use of the Internet is an explanatory variable and that this variable has a mean of 40 percent in the survey and a mean of 25 percent in the census. If this variation derives from differences in the survey instrument or in the implementation (say, one source included Internet use in public locations, but the other did not), then the inclusion of such a variable would lead to a bias in the estimated levels of poverty. If the variation is associated with sampling, this would be less of a concern. However, there is a trade-off between the low R squared in the models and the bias because of nonmatching distributions. In 2012, stage 0 was implemented according to the standard procedure, with only variables similarly defined and with similar distributions included in the regressions. Perhaps because of the stage 0 restrictions, the R2 levels in 2012 were lower than those in 2002 for a similar number of variables. The R2 levels in 2002 were in the range of 0.56–0.65, while the range was 0.46–0.57 in 2012 for a similar number of explanatory variables. The lower levels of R2 in 2012 will lead to larger standard errors in the estimated poverty levels; however, the estimated results will pertain to the year of the survey, and there is no doubt about the potential bias in variables with diverging means. Also, in 2002, to produce a complete dataset, a few variables with missing data were imputed based on a model. This was not an issue in 2012. Stage 1: Estimation models In 2002, six models were established based on the stratification used for sampling in the 2002 household survey (central urban, central rural, mountain, Tirana, coastal urban, and coastal rural) (Betti, Neri, and Ballini 2003). In 2012, only four models were established because the urban and rural strata were combined in the coastal and central areas. Some municipalities and communes have both urban and rural observations, and, for these areas, the estimated headcounts rely on two separate prediction models. In this case, the estimated standard errors of the Foster-Greer- Thorbecke (1984) family of poverty measures and inequality measures are only correct under the assumption that the two estimation models are uncorrelated. This assumption does not seem likely to be true; so, joint models were preferred over split models. Using a combined model could lead to a bias if a single model cannot embed both rural and urban patterns. This does not seem to be a problem because there are no systematic differences between the relative size of the location effect between the 2002 and 2012 maps. If a joint model were not able to capture both rural and urban patterns, one should expect a larger location effect in the joint model, compared with each of the urban and rural models. This is not the case. In 2002, household weights were only applied to the model in stratum 3, but not to other strata. In 2012, household weights were applied to all models. In both years, the models were relatively parsimonious with 10–15 household-level variables and 2–10 enumeration area–level variables. Portraits of Poverty and Inequality in Albania | Annexes 91 Stage 2: Simulation stage At the simulation stage, both sets of maps utilize the ELL approach in simulating the beta coefficients from the main model. The maps vary slightly in the simulation of the cluster and household error terms because the 2002 maps bootstrap from the actual error terms in the estimation model in the survey. The 2012 maps use the normal mixtures approach, an alternative method not developed until 2014, whereby the error terms from the survey models were used to approximate a nonparametric distribution, which was then used in the bootstrapping. Boundaries over time There was only a limited change in the administrative boundaries between the 2001 and 2011 census. Only a single commune was merged. Estimated commune poverty levels It is key that any interpretation of the trends in poverty between 2001 and 2012 at the commune level should only be carried out based on an in-depth analysis of all the data. Such an analysis is beyond the scope of this report, but a few observations and illustrations highlight one aspect. Figure E.1 illustrates the changes in poverty across communes in each prefecture. It shows that there is notable variation in the levels of poverty reduction, mostly within prefectures. An analysis of these changes should take into account that communes are small units for the estimation of poverty. Most poverty maps use areas with larger populations in each area, and, as a reflection of the small populations, the estimates are also associated with notable standard errors. This is illustrated in figure E.2, which shows larger changes in poverty in communes with smaller populations. Hence, much of the large shifts observed in poverty may merely be variations within the standard errors and not necessarily significant changes in poverty. 92 Portraits of Poverty and Inequality in Albania | Annexes Figure E.1 Change in Commune Poverty Rates over Time, by Prefecture 80 60 40 20 0 -20 1 10 11 12 2 3 4 5 6 7 8 9 Figure E.2 Absolute Change in Poverty over Time and Log Number of Households, Communes 80 60 40 20 0 4 6 8 10 12 Log number of HHs in commune in 2012 Portraits of Poverty and Inequality in Albania | Annexes 93