SETTLEMENT

Settlement and household survey report SETTLEMENT AND HOUSEHOLD SURVEY REPORT 0 Settlement and household survey report Introduction Chapter 1 – Sampling methodology 1.1 Sample size 1.2 First stage sampling: food agents 1.3 Second stage sampling: households 1.4 The questionnaire 3 7 8 10 11 11 Chapter 2 – Settlements’ population Chapter 3 - Settlements: needs in services and infrastructures 3.1 Settlement housing 3.2 Education 3.3 Health care 3.4 Social and community services 3.5 Agricultural services in the rural settlements 3.6 Water, electricity, sewage and roads 13 19 19 20 22 24 26 26 Chapter 4 – The demographic structure of the household 4.1 Age structure and dependency ratio 4.2 Demographic strength of households 4.3 IDP families 29 29 32 33 Chapter 5 – The literacy skills Chapter 6 – The wealth of households 6.1 The employment structure 6.2 The livelihood of the household 35 37 37 40 Chapter 7 – Housing conditions 7.1 Buildings 7.2 Dwellings 7.3 Housing conditions 7.4 Adequacy of housing 7.5 Expectations 43 43 45 46 47 51 Chapter 8 - Urban - Rural – Collective towns 8.1 Household‟s profiles 8.2 Housing conditions 53 53 54 Chapter 9 - Housing vulnerability 9.1 Economic profiles of the household 9.2 Demographic profiles of the household 9.3 Vulnerable households 56 57 64 70 Chapter 10 - Final recommendations 10.1 Policy review 73 75 Appendix A: 79 1 Settlement and household survey report S&H Survey Team Appendix B 79 80 2 Settlement and household survey report Introduction Within the framework of Habitat‟s activities in Northern Iraq it is of fundamental importance to collect information about the socioeconomic conditions of this area. This information is needed for policy purposes: (i) to improve the implementation of humanitarian programs related to local population needs and therefore to a more adequate identification of target population, (ii) for higher understanding and evaluation of possible impacts of humanitarian action on overall socioeconomic conditions of this area, (iii) for a more precise definition of social domains and geographical areas where to prioritize Habitat‟s project allocation and to measure policy performance. In recent years Habitat has developed many activities that have assisted the Agency in defining its delivery system (turnkey houses, shelter packages, infrastructure and watsan projects and other forms of structural intervention). The single items of this delivery system were defined on the basis of the community work developed by field teams, that were constantly in contact with the major stakeholders of local population such as local representatives and specific social profiles like widow headed families, IDPs, refugees, poor urban people. During the year 2000 a survey was carried out on IDP sites and households, it was the first attempt of addressing the problem in target population analysis and definition in a more „scientific‟ way. As well as this survey other studies were started (rural area, identity groups and housing typologies analyses). With the „Household and Settlement Survey‟ Habitat wants to continue this more systematic approach based on surveying methods and techniques, aimed at producing a first important step toward the setting up of a permanent monitoring system that will answer questions posed by policy makers. Such questions may be related to the following areas of policy concern: - what proportion of local population can be considered vulnerable and therefore in need of external assistance within the framework of humanitarian programs; - in this context how should a housing policy effect local population living conditions in an area where housing problems can not be easily identified, in the sense that problems like lack of housing (homelessness) or 3 Settlement and household survey report - - - substandard housing are probably less visible and thus understandable as in other poor areas of the world; how can the housing problem be related to other typical problems of this area such as low level of activity, fall in traditional agricultural activities and general household vulnerability that is most likely to affect a whole generation of local population in terms of loss of farming (and related) skills; how can resettlement programs that are now in progress help to improve the living conditions of local population not only by increasing the existing housing stock, but also by effecting local economy through building activities and improving the methods of governance; what criteria and leading ideas can be first developed and then applied in order to understand what could be the optimal spatial distribution in allocating housing and infrastructure projects and what type of criteria can be applied in defining urban and rural area allocations. The Settlement and Household Survey (from now on S&H Survey) is therefore supposed to collect information that will help Habitat in answering a very wide range of possible questions of policy and planning concerns with reference either to settlements structure (first part of the survey) or to households and population features (second part). The S&H Survey was planned to be completed in the first half of the year 2001. Instead the work started at the beginning of February with the planning phase, while the data collection phase went from April to the first decade of July, while checks on the accuracy of the data were carried out in the mean time and immediately after. To the data collection phase have participated 40 surveyors (10 in Duhok, 10 in Erbil, and 20 in Sulaimaniyah), 3 supervisors (one for each Field Office), 4 data checkers (1 in Duhok, 1 in Erbil, and 2 in Sulaimaniyah), and 8 data entry (2 in Duhok, 2 in Erbil, and 4 in Sulaimaniyah). The selection of 3 supervisors was necessary to facilitate the implementation and coordination of the survey in the three Field Offices. The 3 supervisors were the focal points to report the activities to the Core Office; according to their reports the logistic/administrative tasks were accomplished through a homogeneous activities‟ coordination. The analyses regarded a quite large range of subjects even though, as far as the part about NI households is concerned, the 4 Settlement and household survey report priority has been given to topics related to housing conditions of North Iraqi population mainly with the aim of describing those social groups that are in greater need of assistance and policy. The first chapter outlines the survey methodology: that is the statistical criteria followed in choosing the sample and the way in which these criteria were implemented in the field when the settlements‟ representatives and the households were interviewed. In chapters 2 and 3 the data about the settlements are analyzed. In the chapter 2 it is described briefly the size of the settlements in terms of number of inhabitants and dwellings, according to the information provided by the Local Authorities, whereas in chapter 3 the study focuses on the needs in services and infrastructures (schools, hospitals, roads etc.), being the minimum unit of observation and analysis each of the three typologies of settlements (urban, rural or collective towns) in each governorate. All of the following chapters of the report concentrate on data produced by the second part of the S&H Survey, namely the household questionnaire. The single parts of the analytical path explained in each chapter are defined in such a way that they approach the final objective of this research, that is to produce a coherent picture of housing conditions and in particular reliable estimates of vulnerable groups in this area. The chapter 4 analyzes demographic structure of the area: the average household and family size, population divided by age and sex, the incidence of vulnerable people (children, elderly, disables) that are taken into consideration with the objective to provide a fair enough estimate (indicator) of household vulnerability from a demographic viewpoint so as to distinguish between strong household‟s structure and weak household‟s structures. The chapter 5 analyses the level of literacy skills of the members of the households. In a population so young as is the one of Northern Iraq more than a quarter of the individuals are of a schooling age. It is important to know how many of them are attending school and how many for one reason or another choose to quit studying. In the chapter 6 the study moves to the economic status of the households and their livelihood. In the first part an accurate description is given of working conditions of the sample population, while the second attacks the problem of the household income, looking not only to the sources of income, but also to the patterns of expenditure, level of family assets and type of financial assistance supplied by local authorities, international organizations or others. 5 Settlement and household survey report The information about the economic status combined with the information about the demographic structure and the literacy skills of households allow us to estimate the level of resilience of the household and to appraise their coping capacity against the hazard events that they may become exposed to. In the chapter 7 the housing conditions of the families are analyzed. Therefore, our attention will be focused on the typologies and dimensions of the dwellings, the standard and the quality of the available services and the level of overcrowding of the housing units. The purpose of this investigation is to provide a first estimate of the number of people living in inadequate (sub-standard) housing conditions. In the chapter 8 we address the problem of different social and housing profiles of the three groups of population: those that are settled in urban areas, those that are living in collective towns and those that are living in rural territories. The poverty line in urban, collective or rural areas may change because of the different purchasing power of households‟ income. The same applies to housing conditions (a lack of facilities as electricity or toilet in urban areas is much worse than in rural areas). In this chapter we try to make the reader more aware of this differences and to go over the appearance of greater concentration of housing disruptions in rural areas. In the chapter 9 we tried to get to a synthesis on the matter of housing vulnerability. The population of the sample will be classified in accordance to three major criteria: housing adequacy of the dwelling, demographic strength and economic wealth status of households. This final part is mostly directed to yield some orientate results on important matters of policy concern. Finally, the chapter 10 contains some recommendations and suggestions provided by the light of the results of the survey. 6 Settlement and household survey report Chapter 1 – Sampling methodology The S&H Survey is addressed to households and to settlements. Due to the size of statistical population (655,000 households and more than 6,000 settlements1) a sample methodology had to be applied to both: households and settlements. While household samples were chosen on the basis of a sampling procedure as defined below, settlements were sampled in such a way that only those settlements were chosen where at least one household sample was located. This made the settlement sample decrease in size: large settlements with a consistent number of household samples are all represented, while the smaller the settlement the lower the chance of a household sample being located in and thus the settlement being selected. Besides this a list of urban settlement was produced in order to make sure that all urban settlements would be covered. We defined urban settlements not on the basis of administrative units distribution but simply looking at the socio-economic profile of the settlement. Settlements were classified as urban where urban type activities were found. Thus only 33 urban settlements were identified. Settlements were divided in three categories: urban, collective towns and rural. Due to the applied sampling methods all urban settlements were covered, while collective towns at 30.6%; rural settlements are represented by a very small sample for which very little statistical significance can be claimed. Therefore for rural settlements the collected data can be considered as indicative as they must be completed in the future either using other data sources or by an extension of present survey to a more statistically significant sample of settlements. While this applies to settlement data it is needles to say that household data in rural areas are not affected by problems of statistical significance since they have been selected on a random base. After accessing the sampling methodologies used by other household surveys in this area in recent years by different UN agencies and other institutions, a two-stage cluster random sample was identified and used for the S&H Survey. Sampling was designed to be carried out on WFP lists of population and food agents (WFP Food Ration database). These were considered, at the moment of launching the survey, the most unbiased lists that could be used for 1 These are the most recent estimates based on the North Iraq Gazetteer supplied by UNOHCI and JHIS. 7 Settlement and household survey report our sampling problem, since the objective of the survey team was to produce a sample that would be as much random as possible, in order to avoid problems of over or under representation of specific categories within the social structure of NI population. At the first stage of cluster sampling, a list of 640 WFP food agents was produced according to the sample size defined (see below). As stated by WFP service standards the food agents are requested to be scattered throughout the territory since they should be accessible to local population within an area of walking distance. This does not mean that the food agents are evenly distributed in spatial terms. They are primarily concentrated in urban areas or in large settlements. To avoid the sample being more representative of urban household categories a second stage of sampling was introduced. Within each cluster or food agent, a number of 5 household‟s samples were randomly selected from a beneficiaries‟ list supplied monthly to the agents by WFP administration. Thus, the number of housing units sampled amounted to 3.200 (about 0,5% of the total population). Since not all of the listed households in the food agents lists necessarily live in the vicinity of the food agent‟s location (some of them may live in small villages close to urban areas) in some cases samples, that were located in other areas (or villages), might have been selected. This procedure made the sample more evenly distributed in spatial terms. The determination of the sample size and the sampling itself were performed separately for each of the statistical areas in which was divided the statistical universe: Darbandikhan, Dohuk, Erbil and Sulaimaniyah. 1.1 Sample size As a first step in defining the size of the sample, a minimum sample size was determined. This minimum size had to ensure the sample to be enough representative to provide significant estimates for each statistical area: for this purpose two different statistical tests were applied. The present being a multi-purpose survey, in the following evaluation tests some target parameters were used, which are supposed to be significant for the distribution of some basic features of NI population. The first evaluation test was based on a proportion, which is an average value for the whole NI region. According to most recent WFP nutrition estimates, on average, in NI 65% of family‟s income 8 Settlement and household survey report is spent for food. Assuming this proportion at a 95% level of confidence, accepting a 5% sampling error and finally assuming 2 as design effect the minimum sample size for each statistical area is equal to 700 units. The second test is based on separate data for each statistical area and takes into account indicators of the variability of some statistical attributes of our population. Among the available data, we chose the simple distribution of family size2 (according to WFP data) to be applied in the evaluation procedure. The mean and variance of these distributions in the four statistical areas are as reported in table A: TABLE A: Area Erbil Suleimanyah Dohuk Darbandikhan Mean (m) 5.31 5.06 6.32 5.37 Variance (s²) 8.45 7.24 11.2 7.97 Considering, as previously, these values at a 95% level of confidence, accepting a 3% sampling error and assuming 2 as design effect, the sample sizes in the four areas should be at least: 680 in Dohuk, 760 in Erbil, 690 in Sulaimanyah and 660 in Darbandikhan. According to these two evaluation tests, a conclusion was drawn that the minimum sample size in each of the four areas should be 700, except for Erbil where this threshold was based at 760 sampled household, for a survey total sample size of 2,860 households. An analysis of the operative and organizational aspects of the survey - along with a comparison of other sample surveys run previously on the same area - showed the opportunity of increasing the total sample size so as to strengthen the significance of the sample. Therefore, to improve the sample, the total sample size was finally established at 3,200 households. Since there are considerable differences in population sizes in the four statistical areas, the gap between the two estimates (2,860 2 For simplicity in this test we don‟t distinguish between the number of families and households. This simplification can not affect the result, since in each household there is at least one family. 9 Settlement and household survey report and 3,200) was distributed proportionally among the four areas. More precisely, at first a 700 units sample size was assigned to each area, in order to be sure to reach the minimum sample sizes required from previous tests, then the remaining 400 units were distributed on proportional basis. The final estimate of the sample size by statistical area was as follows: TABLE B Area Constant part of sample size Erbil Proportional part of sample size 140 130 90 40 400 Total sample size 840 830 790 740 3,200 700 700 700 700 2,800 Suleimanyah Dohuk Darbandikhan Total 1.2 First stage sampling: food agents In each of the sampled clusters five households were observed. In order to reach the requested number of 3.200 households, a sample of 640 food agents was taken at the first stage. In this way the number of sampled food agents was 7,8% of their total number and, on average, a 5% of the households registered at each food agent was observed. The followings are the final numbers of sampled food agents in the four statistical areas: Dohuk 158, Erbil 168, Suleimanyah 166, Darbandikhan 148. After the sample of food agents was selected, a map of the sampled units was produced in order to graphically represent their distribution throughout the territory and overlay it with other significant maps of this area: first with the identity groups maps where settlements are marked by religion and language of local population (the map has been produced by the anthropological study), second with a spatial representation of farming systems that have been envisaged by the rural area studies. 10 Settlement and household survey report In both cases this type of sample checking has yielded satisfactory results and the sample of food agents was finally approved. 1.3 Second stage sampling: households The five households associated to each food agent were selected directly by surveyors. At the food agent‟s shop, the surveyor consulted the list of food rations beneficiaries. From that list a first household was selected randomly (by reading a banknote serial number). The selected household was the first to be interviewed. The next household to be surveyed was the one at the 5th „closest door‟ in any direction taken by the surveyor. The surveyor had to continue this way, by taking the 5th „closest door‟, until the fifth interview was completed. If the surveyor reached the edge of the settlement before gaining the required number of observations, then the procedure went on from the first household‟s location, but in the opposite direction. By following this procedure the selection of the households to be interviewed was in fact random. It is important to note that the households taken from the list at the food agent shop could be located in a different settlement (i.e. village) from the one the food agent belonged to, but still not too far from it. 1.4 The questionnaire In order to fit all the data needs that were requested the questionnaire was divided in the following six data sheets: - Settlement data sheet A (for urban area and collective towns) and sheet B (for rural settlements) - General household data sheet C with the annexed form D for farmers - Individual data sheet E (for active or retired population) and F (for population under 16). The settlement data sheets (A and B) were designed to give information about housing units and population, to produce estimates of needs of services (education, health, community services, public transport) and infrastructures (water, sanitation, electricity, roads). Finally information was collected on major sources of income, occupation and economic activities. 11 Settlement and household survey report The household data sheet “C” is composed of the following sections: neighborhood (environmental quality, level of services and security), building (size, conservation and construction typology), land (ownership, building density and plot organization), dwelling (size, quality, level of utilities, conservation, maintenance needs), family profile, family budget (income, assets, expenses), participation and level of assistance. The farming unit data sheet “D” was filled up only in the case a farmer was detached within the household. It was composed of the following section: farmyard organization, land tenure, land use and agricultural yield, family assets, farming methods and strategies. Finally in the two individual datasheets (over and under 16) the major emphasis was given to health conditions, educational level, transport demand and, for those in working age, to type and quality of occupation. 12 Settlement and household survey report Chapter 2 – Settlements’ population There is a lack of reliable information about the demographic figures in North Iraq due to the disorder which has recently struck this region and that has made impossible to register the fast changes of the resident population. The last reliable data available at Local Authorities statistical offices refer back to 1987, when the North Iraq was still under GOI control. A fairly large number of surveys that have been carried out since then have always looked at one particular aspect of the region or at one particular location (only Villages or only Cities or only Collective towns): any attempt to obtain a reliable overall demographic profile from them failed because of the constant presence of overlapping data, as well as omission. The record of data that best describes the population of Northern Iraq as a whole is the WFP food-ration database3: it refers to January 2001. This data source seems to slightly overestimate the population because of the double registration of some individuals but it is the most trustworthy source of those available and provides a sufficiently accurate representation of the demographic dimension of the inhabitants of the Northern Iraq. TABLE C Statistical area WFP Population Permanent population of the settlements sampled in the S&H Survey 641,568 1,048,793 839,537 262,647 2,792,545 % of sampled population 83.8% 82.6% 71.6% 77.3% 78.7% Duhok Erbil Sulaimaniyah Darbandikhan NORT IRAQ 765,483 1,269,115 1,172,303 339,951 3,546,852 Trying to provide the exact number of the whole Northern Iraq population goes over the intention and the possibilities of this survey. What we can ascertain from the data of this survey is whether there 3 This database is constantly updated. For more details see the report of Marco Carbonara, North Iraqi demographic data, June 2001, Habitat-Erbil. 13 Settlement and household survey report have been significant changes in the population size in the last decade and, if these changes exist, evaluate their magnitude and distribution between the statistical areas. Besides that, a demographic distribution of the individuals according to their classes of ages will be provided, firstly considering the whole population and then the population divided by sex. It has to be said that the 33 urban settlements (table D) covered by the survey represent the total universe of urban areas under observation (as defined above) whereas the rural settlements and collective towns are two samples (covering around the 30% of the collective towns4 and some 3% of the many villages scattered in NI5). TABLE D: Governorate Duhok Erbil Sulaimaniyah Darbandikhan NORTH IRAQ Urban 6 12 10 5 33 430 600 530 355 1915 CT 26 12 11 3 52 205 120 155 145 625 Rural 21 22 26 44 113 155 120 145 240 660 In table E we represented the geographic distribution of the selected settlements. This map has been overlaid to other information available to Survey team such as rural areas categories and ethnic groups distribution in order to ascertain that the samples would not be biased by an uneven spatial distribution of selected samples. The table 2.1.1 represents the population of the current sampled settlements now and those from 1990, in the four statistical areas, divided by urban, rural and collective towns. During the two years under examination the Northern Iraq number of residents grew of 15.4%: there was an increase of 30.4% in the urban areas, an increase of 145.4% in the rural areas, and a decrease of 31.8% in the collective towns. In the Duhok‟s governorate the population counts today 427,033 inhabitants in the urban area, 171,171 in the collective towns and 43,364 in the rural area. In the 1990 the demographic figures stood at 332,981, 130,756, and 28,935 in the three areas, respectively. 4 5 See the IDP site and family report, Erbil, January 2001 Estimates based on the Gazetteer produced by UNOHCI and JHIS. 14 Settlement and household survey report Duhok went from 143,000 to 224,500 inhabitants (57%), Zakho from 148,581 to 148,681 (0.07%), Summail from 16,580 to 23,670 (42.8%) and Akre from 18,720 to 23,198 (23.9%). TABLE E In the Erbil‟s governorate we counted 927,047 inhabitants in the urban centers, 91,854 in the collective towns and 29,838 in the rural area. In the 1990 there were 667,193, 113,493, and 20,215 inhabitants in the three areas, respectively. Benslawa went from 28,000 to 45,000 inhabitants (60.7%), Soran/Diana from 20,868 to 42,100 (101.7%), Koysinjaq from 30,616 to 34,415 (12.4%), and Daratu from 24,565 to 32,166 (30.9%). In the urban centers of Sulaimanyah‟s governorate the population reaches 661,552 individuals, in the collective towns 134,404, and in the rural area 43,581. In 1990 the same zones counted 517,052, 361,909, and 1,720 residents. Sulaimaniyah passed from 453,326 to 436,146 inhabitants (-3.8%), Hallabja from 0 – it was destroyed in 1988 and successively rebuilt - to 46,400, Raniyah 15 Settlement and household survey report from 25,446 to 45,000 (76.8%), Qalladiza from 0 – it followed the same misfortune of Hallabja - to 40,646. Finally, in the Darbandikhan area, the data of the sample show 168,050, 71,423, and 23,174 inhabitants now and 157,440, 81,070, and 6,164 in 1990. Kalar went from 68,000 to 70,564 inhabitants (3.8%), Chamchmal from 49,200 to 53,850 (9.5%), Shorsh from 42,720 to 33,145 (-22.4%). The population grew and keep on growing largely in all the Northern Iraq (this fact is likely to be explained with the high birthrate, the relative peaceful period the country is going through and the consequent flourishing of economic activities that draw people from other parts of the country) even though the phenomenon assumes remarkable differentiations between one zone to another. In the governorate of Duhok, located in a very strategic point between Turkey, Syria, and the rest of Iraq, in the years between 1990 and 2001 there was a population increase of 28.3% in the urban centers, of 30.9% in the collective town (the collective towns located on the border with Turkey have been attracting a large number of people dedicated to profitable trading activities) and of 50% in the rural centers (in some of them, like Sarsink, the flow of IDPs keeps on increasing the number of residents, some other, of smaller dimension, like Bardarash, are the gathering point and the base of smugglers). In the governorate of Erbil the number of residents in the urban areas increased by 39%, that of residents of collective towns decreased of 19% (the new IDPs arriving from the area around Kirkuk are less than those who, taking advantage of rehabilitation process in progress, are moving back to their original place) and that of rural areas increased by 47.6%. In the governorate of Sulaimaniyah the population grew by 27.9% in the urban areas, fell by 62.9% in the collective towns and gained more than 2000% in the rural areas. These apparently surprising figures can easily be explained considering the history of the past years. The governorate of Sulaimaniyah was one of the areas that suffered the worst consequences of the war and internal conflicts in the 80s and in the early 90s. Many rural villages were destroyed and their inhabitants forced to move to collective towns near Erbil or Sulaimaniyah. From 1991 the political situation has changed and the people who had been displaced have been able to return to the original settlements: the rural areas, deserted in the 1990, have been reoccupied by the former dwellers. 16 Settlement and household survey report The same phenomenon occurred in the Darbandikhan Area, even though the dimensions of the demographic movement have been of quite smaller proportion, according to the demographic consistency of that zone. In some areas IDPs from Kirkuk continue to arrive, but their number is much lower than the number of IDPs that go back to the formerly inhabited settlements. The presence of temporary population – mainly displaced people forced to move to urban centers and that for some months in the year go to cultivate land in their place of origin - is very high in the urban and rural areas and practically absent in the collective towns (the collective towns of the governorate of Erbil constitute an exception). The table 2.1.2, which shows a comparison between the temporary and the permanent population in the Northern Iraq wholly considered, indicates that the ratio of the two figures is of 10.6% in the urban areas, of 3.1% in the collective towns, and of 7.3% in the rural areas. In the governorate of Duhok, only the urban centers stress a noteworthy presence of temporary people (4.0%). In the urban zones of Erbil there are about 9 temporary people for every 100 of non temporary, in the urban areas of Sulaimaniyah 18 whereas in the collective towns of the governorate of Erbil there are about 13 temporary residents to 100 of permanent residents. The data point out the biggest ratio in the rural area of Erbil (19 temporary to 100 permanent), followed by those of Darbandikhan and Sulaimaniyah. Two kinds of reasons could have provoked the temporary residence of these people: the first reason is related to safety issues, the second one to financial situations. The ones that reside for safety issues are individuals who come either from other areas of the same governorate or from other parts of Iraq and that are waiting for the political situation to improve in order to return to the places from where they were forced to move. The situation of the people who live temporarily in the rural areas (as previously seen, the figures indicate that the phenomenon assumes considerable dimension in the governorate of Erbil and in the Darbandikhan area) can be regarded differently. Great part of these people are farmers who move seasonally - mainly in the summer - to their fields to cultivate the land and that go back to urban residences to devote themselves to other activities when the work in the country is over. A comparison between the percentage of the male and female population reveals a remarkable predominance of the women in the 17 Settlement and household survey report collective towns located in the governorate of Erbil and in the Darbandikhan area, in the urban centers of the governorates of Duhok and Sulaimaniyah, and in the rural area of the governorate of Sulaimaniyah. The men outnumber the women in the rural area of Erbil. In the remaining area there is an overall balance between the demographic figures of the two sexes. 18 Settlement and household survey report Chapter 3 - Settlements: needs in services and infrastructures In this chapter we analyze the quantity and the quality of the available services in the four territorial zones of North Iraq. Firstly our attention will briefly go over the number of dwellings in the settlements and then concentrate on the schools, health centers, social and community services, roads and utilities, evaluating the adequacy of each service according to the requirements of the population. 3.1 Settlement housing On average, in the urban area, a dwelling is inhabitated by 5 (in the governorate of Erbil) or 6 (in the governorates of Sulaimaniyah and in Darbandikhan Area) people. In the governorate of Duhok the number of people per dwelling is bigger (9). The average household size is greater in the governorate of Duhok (10 persons per dwelling in the urban settlements and 7 in the collective towns) and smaller in the governorate of Erbil (around 5 people in all the three types of settlement) whereas in the urban towns of Sulaimaniyah and in the CT of Darbandikhan Area it is about of 6 – 7 people. In the rural areas of Northern Iraq a dwelling is inhabited by 6 persons, with the exception of the rural area of the governorate of Duhok, where, on average, 9.2 people live in each dwelling. The number of deserted houses is generally irrelevant to the total number of the houses. This is due in part to the difficulty to collect sufficient information about the not inhabited dwelling, in part because if a house in discrete conditions is abandoned, it doesn‟t remain deserted for long: a group of people take soon possession of it. The principal reason for a house being deserted is that the owners are moving to another place because of economic problems: the abandonment of the dwelling is often temporarily and concerns in most of the cases houses of farmers who leave their residence for some periods of the year in order to go to work in their fields, far from their principal home. 19 Settlement and household survey report 3.2 Education The number of students exceeds the 24% of the permanent population in all the four areas of the Northern Iraq (see table 3.2.1) and the number of pupils enrolled in the primary school is generally high: very few are the settlements – located in the most isolated rural areas - where the percentage of pupils attending school doesn‟t reach the 50%. In the urban areas the number of people attending school is higher than in the rest of the territories: 29.8% in the governorate of Duhok, 23.9% in the governorate of Erbil, 27.1% in the governorate of Sulaimaniyah, and 27.7% in the Darbandikhan Area. Considering the Northern Iraq as a whole, the ratio between number of student and permanent population reaches the 26.3% in the urban areas, the 24.7% in the collective towns, and the 23.6% in the rural areas. Often the number and the quality of available schools and connected structures are not enough to satisfy such a great school attendance, especially in the rural areas and in some small collective towns, where the public interventions are less frequent than elsewhere. Here are some significant figures. In the urban centers for every 1,582 inhabitants there is a school in the governorate of Duhok, one for every 2,088 in the governorate of Erbil, one for every 1,929 in the governorate of Sulaimaniyah, and one for every 1,514 in the Darbandikhan Area. The situation gets worse in the collective towns where there is a school about every 2,000 residents but improve, at least respect to the ratio between number of schools and number of residents, in the rural area where there is one school every 600 – 700 people (see table 3.2.2). The ratio of the number of students to the number of available classrooms provide a good synthetic index of the quality of the school system in terms of available space inside the buildings. Considering as adequate one classroom for every 30 students (every 35 in urban settlements with high demographic density, every 20 in rural areas), from the data of the sample it appears clear that the real situation is far from this standard. In all the sampled areas the space put at students‟ disposal is very scarce. The peak is reached in the cities of the governorate of Sulaimaniyah where there is a classroom every 67 students, but the space problems involve almost all the schools of the region (see table 3.2.3), in particular the primary schools where students take lessons in double or triple turns so everyone is able to attend classes. The settlements that most lack in classrooms are located in the urban areas (a rough estimation indicates that Sulaimaniyah need 20 Settlement and household survey report more than 2,000 classrooms, Erbil about 600, Duhok about 300 and Darbandikhan about 190) but the necessity of space where to teach is pressing everywhere. These space problems do not concern only the number of classrooms but also the presence and the adequacy of sport facilities, multi purpose halls and laboratories, in which all the school buildings seem to be extremely lacking: in the collective towns and in the rural areas is a fortune to find a school that dispose of a gym or of a room where to carry out experiments, researches and practical lessons. Besides the space problem many schools present a lack of basic services. Two are the main reasons of this shortage: a general insufficiency of the public supply systems and a intrinsic decay of the school buildings due to their old age and the insufficient maintenance. It has only to be said that only less than 25% of the sampled schools were built after the 1997. Altogether about 64% of the schools have toilets. In the collective towns of Erbil (53.7% of the school buildings), Sulaimaniyah (33.4%) and Darbandikhan (46.2%) the situation is critical whereas in the urban areas the hygienic conditions are higher (61.5% of the schools located in the Darbandikhan Area, almost all of those of the governorate of Duhok, 58.5% of those of Erbil, and 51.6% of those of Sulaimaniyah dispose of lavatories). In 40-50% of the schools of the collective towns and of the rural areas (with the exception of those located in the governorate of Duhok) there is no water inside the buildings and in most of the schools of the governorate of Erbil the electric energy is very rare (the electric supply system of the statistical areas of Duhok, Darbandikhan and Sulaimaniyah seems to work definitely better: almost all the schools of the urban areas and of the collective towns benefit of electricity for at least many hours during the day). A great percentage of the school buildings have maintenance problems. In the Durbandikhan Area 53.8% of the schools located in the collective towns, 32% of those located in rural areas and 13.8% of those of urban centers have a structure that has either to be repaired or to be rebuilt. In the governorate of Duhok these percentages are of 21.6%, 24.3%, and 34.4%, respectively in the urban, collective towns, and rural areas, whereas in the governorates of Erbil and Sulaimaniyah the percentages become of 27.5%, 48.8% and 63.6% and of 36.6%, 58.3% and 37.8%. Most of the schools, especially high schools, technical schools and institutes, are located in the high-density populated urban centers of the region, where the majority of students live and where the 21 Settlement and household survey report demand for education services is higher. Generally in the outskirts only kindergartens, primary and sometimes intermediate schools are available. The three universities of the region are located in the cities of Erbil, Sulaimaniyah and Duhok. 3.3 Health care The number of medical facilities and their size, by itself, seems to be adequate to meet the needs of the whole population even though the quality of these services in most of the cases is at sub-standard level. Hospitals and health centers are spread out across all the areas of the North Iraq where it has been estimated that there is a hospital or health center for every 14,755 people in the urban areas, one for every 9,377 people in the collective towns, and one for every 3,110 people in the rural areas (see table 3.3.1). In the governorate of Duhok there is one hospital or health center in use for every 10,168 residents in the urban area, one for every 7,781 in the collective town, and one for every 5,421 in the rural area. In the governorate of Erbil the proportion is one health center for every 18,919 citizens in the urban area, one for every 8,350 in the collective town, and one for every 2,487 in the rural area whereas in the governorate of Sulaimaniyah the proportion is one for every 14,075, one for every 11,200, and one for every 2,723 in the three areas under observation and in the Darbandikhan Area one for every 16,805, one for every 14,284, and one for every 2,575. The number of medical staff working in the hospitals and health centers seems to be adequate, too. On average, in the whole Northern Iraq, there is a doctor or a member of the paramedical staff every 244 people in the urban areas, every 529 people in the collective towns, and every 419 people in the rural areas. In the governorate of Duhok these are the ratios between permanent population and medical staff: 278 in the urban areas, 641 in the collective towns, and 647 in the rural centers. In the governorate of Erbil (the biggest hospital of the Northern Iraq is located in the main town) the figures of the three respective zones are 308, 356, and 321, in the governorate of Sulaimaniyah 163, 460, and 321 and in the Darbandikhan Area 462, 1,035 and 610 (see table 3.3.2) If the number of hospitals, health centers, doctors, nurses and other hospital staff is sufficient to grant adequate medical assistance to all the inhabitants of North Iraq, the quantity and the quality of the 22 Settlement and household survey report structures, services, and equipments whose these health centers and medical personnel are provided with are not up to serve in a constant way the requirements of the people who need care. In particular in the rural areas and in the collective towns far from the big cities the necessity of maintenance and improvement interventions are pressing and the problems connected with the scarcity of satisfactory water, electricity and sewer systems are more urgent than elsewhere. The number of available beds compared with the resident population is one of the significant indicators of such a structural deficiency of the health centers (hospitals and health centers literally called), especially of those located in the sub-urban areas (see table 3.3.3). According to the data provided by WHO, there should be 4 beds for every 1,000 people. The results of the survey show figures that are far from this standard. In the governorate of Duhok there is one available bed for every 425 people in the urban centers, one for every 3,493 in the collective towns and one for every 21,682 in the rural areas. The situation seems a little bit better in the governorates of Erbil and Sulaimaniyah where there is one bed for every 429, 10,206, and 2,295 people in the former and one bed for every 372, 10,339, and 1,557 in the latter, and a little bit worse in the Darbandikhan Area where there is one bed for every 913 inhabitants in the urban area, one bed for every 4,762 in the collective towns, and one bed for every 11,587 in the rural zone. A serious shortage of basic hygienically services connected to a more general lack of water, electricity and sewer frameworks can be found in the health centers located in the most deteriorated parts of the region. In the urban areas most of the health centers have a continuous water supply and a sufficient number of toilets (from this point of view the urban settlements of Darbandikhan Area and the governorate of Sulaimaniyah are those that present the worse conditions), whereas in collective towns 70% of health centers have water (40% in the Darbandikhan Area, 100% in Duhok, 75% in Erbil and 50% in Sulaimaniyah) and about 50% toilets and in the rural areas 75% have an adequate water system and about 75% toilets. The complete inadequacy of a public electricity system in the whole North Iraq (with the exclusion of small zones of the governorate of Sulaimaniyah and of the Darbandikhan Area where are located the electrical power plants that produce the major quantity of the electrical energy, near the lakes of Dukan and Darbandikhan) deprive most of the health centers (but not only them, even the other public buildings and the private houses) of a permanent electricity 23 Settlement and household survey report supply and force them to use public or private generators. The data of the sample concerning the availability of permanent electricity supply are significant. In the governorate of Duhok 7% of the health centers of the urban areas, none of those of the collective towns, and none of those located in rural areas have a continuous supply of electricity in any way. In the governorate of Erbil the percentages become of 14.9%, of 0%, and of 7% in the three geographical areas, in the governorate of Sulaimaniyah of 12.2%, 0%, and 17.6%, and in the Darbandikhan Area of 45.5%, 60%, and 0%. Most of the times private generators are necessary to overcome this lack of electricity supply: on average, 87.3% of the hospitals and health centers declared to use them. Only about 30% of the hospitals and health centers buildings were built recently (after the 1997): 60% of those located in urban zone of the Darbandikhan Area and 45.5% of those located in the urban zone of the governorate of Erbil are the highest percentages. Due to their old age much maintenance work is needed to structure, at the roof, etc. Other common problems regard the available useful space (e.g. the number of dormitories) that often is not sufficient to give hospitality to all the in-patients (it is a very widespread problem in the collective towns where the small dimensioned health centers are constituted, in the best cases, by two or three rooms, but that includes more than half of the other hospital buildings located in the urban and rural areas), the number of specialized wards and laboratories in which tests, analyses and research are carried out, adequate medical equipment and medicines whose lack is surveyed in both the big hospital of the cities and the small health centers of the suburbs. Finally, it must be kept in mind that as well as the hospitals and health centers there are a valid number of clinics, doctor assistant ambulatories and prenatal care centers (one for 2,800 people in the urban areas, one for 5,649 in the collective towns, and for 3,999 in the rural areas) where the deficiency of beds, structures and equipments is even more relevant than in the bigger hospital centers (see tables 3.3.4, 3.3.5 and 3.3.6). 3.4 Social and community services Little has been said about the social and community services because of the insignificant number of structures sampled in the survey. Libraries, public computer centers, gyms, swimming pools, 24 Settlement and household survey report tennis courts, basket fields, cinema and theatres seem all to be concentrated in Sulaimaniyah, that from the point of view of leisure and entertainment is the most lively city of Northern Iraq and the center that offers the highest number of attractions and amusements. Darbandikhan and Duhok are the other two cities in which some sports and cultural structures are present. Football fields, parks, green areas and common spaces for children (inexpensive structures that need of less frequent maintenance works) are scattered over the whole country. More interesting is the information concerning the presence of markets (temporary open market, permanent open market, retail market, supermarket, wholesale market, and animals market) in the settlements, from which it is possible to form an idea of the level of economic activities and the consequent development possibilities of the different areas under observation. On average, in Northern Iraq, it has been surveyed a market for every 103 people in the urban areas, one for every 151 people in the collective towns and one for every 263 people in the rural areas (tab. 3.4.1). Here the figures of the three considered governorates and of the Darbandikhan Area are following. In Duhok there is one market every 110 people in the urban area, one for every 75 people in the collective towns and one for every 137 people in the rural areas. In Erbil the numbers are 88, 919, and 368 whereas in Sulaimaniyah and in Darbandikhan are 115, 264, 907 and 173, 343, and 266, respectively. The governorate of Duhok is undoubtedly the economical richest area of North Iraq. As we have already mentioned the most fervent business activities take place here, mainly in Duhok, that is the center around which all the principal business activities gravitate, and in the settlements near the border with Turkey and the GOI. Duhok, Zakho (on the border with Turkey and Syria is the city that produce, collecting taxes on the goods that transit in and out of the governorate, the highest income of Northern Iraq), Akre (a district more important for its demographic dimension and for political matter than for its geographical position), Sumel (near the border with GOI), besides some collective towns like Kalakchi and Qasrok collective, are the centers where the biggest number of markets are concentrated. It is important to notice that the only two supermarket of the Northern Iraq are situated in this governorate, the first one in Duhok and the second one in Zakho. 25 Settlement and household survey report Even in the other statistical areas the major trading activities (and consequentially the biggest number of exchange of goods points) are located in the most geographically strategic areas: Raniyah, Qalladiza and Hallabja (all of them near the border with Iran) in the governorate of Sulaimaniyah and Darbandikhan, Kalar, Kifri and Takia (these last two near the border with GOI) in the Darbandikhan Area. 3.5 Agricultural services in the rural settlements In the table 3.5.1 the distribution of the working population over total rural population by type of job is represented. 10.3% of the total population are farmers whereas almost 4% have other jobs. The percentage of farmers is of 7.2% in the governorate of Duhok, 14.2% in the governorate of Erbil, 9.8% in the governorate of Sulaimaniyah, and 12.3% in the Darbandikhan Area. The available services, comprising cereal, beans, vegetable, and fruit storages, tools storages, garages, cattle sheds, storages tank for irrigation, for who cultivates land are very scarce (see table 3.5.2). Only 11.5% of the sample declared to have one of the abovementioned services, not presenting the figures in the rural areas of the different governorates considerable differences between each other (9.5% in Duhok, 9.1% in Erbil, 15.4% in Sulaimaniyah, and 11.4% in Darbandikhan). For the farmers it is almost impossible to purchase agricultural tools in the settlement in which reside or in the nearby villages: the only way is to reach the closest urban center provided with specialized shops (tab. 3.5.3). Even the settlements provided with working irrigation channels or veterinary centers constitute the minority of the sample (tables 3.5.4, 3.5.5 and 3.5.6). The former are present in 23.9% of the rural areas of the whole Northern Iraq (23.8% in Duhok, 22.7% in Erbil, 26.9% in Sulaimaniyah, and 22.7% in Darbandikhan) whereas the latter in 11.5% (28.6% in Duhok, 4.6% in Erbil, 15.4% in Sulaimaniyah, and 4.6% in Darbandikhan). 3.6 Water, electricity, sewage and roads The analysis carried out on the working water, electricity and sewage supply systems compared to the whole existing network seems to produce, in some statistical areas, paradoxical outcomes, showing a better situation in those zones (collective towns and rural 26 Settlement and household survey report areas) where one should reasonably expect to find the less adequate conditions: the highest percentages of working network are registered in the outskirts instead of in the urban centers where the public work interventions are undoubtedly more frequent and of more relevant importance. This apparent inconsistency can be explained with three reasons: 1) a lack of information about the real dimension of the network from the local authorities interviewed in the survey, 2) the trifling water, electricity, and sewage supply system of the collective towns and rural areas compared to that present in the urban centers, 3) the inclination to consider as the total network kilometers only the kilometers of working network. Anyway, beyond of these incomplete and patchy data, it is possible to explain correctly some general results. In the whole urban areas 77.1% of the water supply system is in function (see table 3.6.1). This quite low ratio is, by itself, worrying but assumes a more negative value if compared with the rest of the sample (85.8% in the collective towns and 72.1% in the rural areas). In the urban areas of the governorate of Erbil (71.4%) and Sulaimaniyah (66.2%) there are the lowest percentages of working water supply system whereas in the governorate of Duhok (99.6%) and in the Darbandikhan Area (86.4%) the situation appears to be much more acceptable. In the collective towns the worst situation is found in the Darbandikhan Area (25.4% of the water system is working) and among the rural areas the Sulaimaniyah‟s zone shows the lowest percentage (42.9%). The percentages of working electricity supply system are of 70.3% in the urban areas, 89.5% in the collective towns, and 79.7% in the rural areas (see table 3.6.2). The governorate of Duhok seems to have a perfectly working electricity supply system, the Darbandikhan Area a satisfactory level of adequacy whereas the worst conditions are registered in the governorate of Erbil (the percentage of electricity system in function is of 26.6% in urban area, 20% in CT and 20.2% in rural area). The governorate of Sulaimaniyah reflects quite well the average profile of the whole North Iraq (68.7% in urban settlements, 90.5% in CT and 72% in rural areas). The extremely scarce sewage supply system (especially in the CT and in the rural areas) doesn‟t permit a judgment on its functionality. Exactly the sewage system is considered completely inadequate by 70% of the interviewed urban centers (supply 28.8% of the population), 92.3% of the collective towns (supply 6% of the population), and 99.1% of the rural settlements. Only in the urban 27 Settlement and household survey report areas of the governorate of Sulaimaniyah and the Darbandikhan Area the conditions appear a little bit better. The water supply system is considered inadequate by 27.3% of the urban sample (where 84.7% of the resident population benefits by it), 25% of the collective towns‟ sample (where 64.2% is supplied), and 42.5% of the rural sample (see tables 3.6.4 and 3.6.7). The most alarming situation is found in Sulaimaniyah (50% of the urban settlements, 72.7% of the collective towns, and 61.5% of the rural settlements express an absolutely negative judgment) and in the Darbandikhan Area (here the percentages of the survey are of 40%, 33.3%, and 52.3%). Finally, the 54.6% of the urban settlements (where 58.7% of the people are supplied), the 48.1% of the collective towns (where 85.4% of the people are supplied), and the 58.4% of the rural settlements consider the electricity supply system totally inadequate (tab. 3.6.5 and 3.6.8). From the data of the sample it appears clearly that the road network is also insufficient (see tab. 3.6.10): only in 18.2% of the urban settlements (16.7% in Duhok, 25% in Erbil, 20% in Sulaimaniyah, and 0% in Darbandikhan) and in the 3.9% of the collective towns (the 8% in Duhok and the 0% in Erbil, Sulaimaniyah, and Darbandikhan) it meets the necessary requirements of functionality and adequacy. The upkeep of the roads is very rare and consequentially the quality of the roadbed is awful (see tab. 3.6.11): the safety of the driving is compromised. In 24.2% of the sampled urban settlements, in 13.5% of the collective towns, and in 8% of the urban settlements road maintenance works are carried out. The situation results better in the governorate of Duhok and Erbil and catastrophic in Sulaimaniyah and Darbandikhan. As far as the paving is concerned the data confirm the more acceptable condition of the roads located in the governorates of Duhok and Erbil. On average, in the whole Northern Iraq, the paved roads constitute 40.2% of the total roads in the urban areas and 46.1% in the collective towns. In Duhok the percentage are of 47.2% and 66.2%, in Erbil of 57.3% and 60.6%, in Sulaimaniyah of 35.8% and 32.9%, and in Darbandikhan of 35.5% and 26.8%. 28 Settlement and household survey report Chapter 4 – The demographic structure of the household According to the definitions adopted, by household we mean “a basic residential unit where economic production, consumption, inheritance, child rearing, and shelter are organized and implemented” and by family “a residential kin group composed of a woman, her dependent children, and at least one adult male joined through marriage or blood relationship”6 Therefore household and family are two different concepts, the former comprising the latter: all the members of the same family belong to the same household but not all the members of the same household belong to the same family, the household being formed by one or more families. This chapter, in particular, is dedicated to the demographic structure of the household. After providing some general results from the survey about the whole population of Northern Iraq, the study will concentrate on the household-unit in order to evaluate its strength (soundness) in regard to the age, the sex, and the health status of its members and in particular of the household head. It must be kept in mind that the number of households interviewed in the survey was chosen according to a statistical expression that guarantees the significance of the sample in each of the four areas under examination. Due to the application of this expression, the number of sampled households is not proportional to the total number of resident households in each area and to calculate the data relative to the whole NI a weighted average has been used instead of a simple average, in order to reflect the true weight brought by every household to the total sample. 4.1 Age structure and dependency ratio As it can be seen from the graph 4.1.1, representing the percentage distribution by ages of the sampled population of NI, the frequency distribution appears sharply skewed to the right (positively skewed). This graphical representation and two synthetic distribution indexes (the average age equal to a little less than 23 years - with irrelevant differences between the two sexes- and the median age the age that split the population in two parts, each of 50% - of 18 6 William A. Haviland, Cultural Anthropology, Holt Rineheart and Winston Inc, 1990 29 Settlement and household survey report years) indicate that the population is characterized by a strong preponderance of young people. The distribution graph also presents three observable hollows areas, one corresponding to the age class between 6 and 14, one corresponding to the ages between 35 and 42 years and another corresponding to the ages between 50 and 60. The reasons of these anomalies in the frequencies distribution are two. The low number of children aged 6 to 14 is due to the high infantile mortality ratio, caused by the lack of food and medical treatment, that struck the country between 1991 and 1998. The many conflicts that upset the country from 1980 to 1994 and heavily affected at least two generations of people explain the sudden falls of frequencies in the other classes of age. Table F Area DA DU ER SU NI 0-15 Male 1,046 1,539 1,308 1,030 4,874 0-15 TOTAL 16-65 16-65 TOTAL Ovr-65 Ovr-65 Female 0-15 Male Female 16-65 Male Female 1,077 1,444 1,304 1,009 4,772 2,123 2,983 2,612 2,039 9,646 1,307 1,732 1,622 1,444 6,107 1,475 2,782 1,761 3,493 1,681 3,303 1,476 2,920 6,323 12,430 68 93 106 107 390 82 112 103 91 388 TOTAL Tot Tot OVR Total Male Female 65 150 2,421 2,634 5,055 205 3,364 3,317 6,681 209 3,036 3,088 6,124 198 2,581 2,576 5,157 779 11,371 11,483 22,855 The dependency index, given by the ratio between the population that, in a presumptive way, is considered economically dependent (people aged between 0 and 15 and over 65 years) and the population economically independent, is a very significant indicator of the demographic strength of a country (table G). In fact, firstly it measures the amount of individuals that are unable to provide for themselves per every person in working age. Secondly, in a dynamic perspective, the dependency index calculated taking into consideration, as far as the dependent population is regarded, only the youngest (Index1 in table G), provides an estimate of the number of individuals that are going to take over the present working populations and, indirectly, of the future demand for labor. In the entire country the dependency index is equal to 84 (which means that there are 84 dependent people for every 100 independent individuals). Considering the high percentage of young people present in NI, 84 is a value that set NI among the developing countries with high fertility rate (the dependency index generally 30 Settlement and household survey report ranges from 45 in developed countries to 100 in the developing world). Two observations arise from the comparison of the values of the dependency index in the four areas as well as in the two sexes. The first one is that there are remarkable differences of the load of dependent population in the different geographical regions (81.7 in Darbandikhan, 91.3 in Duhok, 85.4 in Erbil and 76.6 in Sulaimaniyah). The differences are due not as much to the load of the oldest class of ages (very low, in percentage terms, compared to the whole population) as to the ratio of the youth with the independent people. This ratio, in fact, indicates that the percentage of the youngest groups are not equally distributed. The dependency index relative to the youngest classes of ages assumes the highest value in the governorate of Duhok (85.4), decreases in the governorate of Erbil (79.1) and in the Darbandikhan Area (76.3), and reaches the lowest level in the governorate of Sulaimaniyah (69.8). The second observation concerns the differences of value found between the two sexes (table G), even in this case more stressed among the ratio between the youngest classes of the population and the independent people than between the elderly and the independent population. In NI (79.8 males against 75.5 females), as well as in each of the four considered regions (80 against 73 in Darbandikhan, 88.9 against 82 in Duhok, 80.6 against 77.6 in Erbil and 71.3 against 68.4 in Sulaimaniyah), the dependency index is sensibly higher for the male population than for the female. Table G Index1 Index1 TOTAL Index2 Index2 TOTAL Index3 Index3 TOTAL Male Female INDEX1 Male Female INDEX2 Male Female INDEX3 80.03 88.86 80.64 71.33 79.8 73.02 82.00 77.57 68.36 75.5 76.31 85.40 79.08 69.83 77.6 5.20 5.37 6.54 7.41 6.4 5.56 6.36 6.13 6.17 6.1 5.39 5.87 6.33 6.78 6.3 85.23 94.23 87.18 78.74 86.2 78.58 88.36 83.70 74.53 81.6 81.70 91.27 85.41 76.61 83.9 Area DA DU ER SU NI 31 Settlement and household survey report 4.2 Demographic strength of households On average the NI household is composed by 7.1 members (with significant differences among the four areas: 6.8 in Darbandikhan, 8.5 in Duhok, 7.3 in Erbil and 6.2 in Sulaimaniyah) and the family by 6.1 members (6.3 in Darbandikhan, 6.9 in Duhok, 6 in Erbil and 5.7 in Sulaimaniyah), of whom 2.6 are young people aged between 0 and 15 years and 3.5 are adults. As far as the marital status of the household members is concerned (see tab. 4.2.1) the data referring to the whole NI indicate that 44.3% of the individuals interviewed in the survey are single (49% of males and 39.8% of females), 50% are married (both the sexes present the same percentage), 5.2% are widowed (0.9% of males and 9.4% of females), 0.22% are divorced (0.1% of males and 0.4% of females) and 0.28% are separated (0.2% and 0.4% in the two sexes). The percentages in the four areas reflect the general data of the country with the exception of the number of widows that in the governorate of Duhok represents 5.3% of the population whereas in all the other three areas represents the 10%. The high percentage of widows (our analysis pays particular attention to widows because they constitute a potentially vulnerable category of people, towards which the Habitat resettlement and rehabilitation policy must concentrate its attention) is another negative result of the wars and internal conflicts that upset this area. In Duhok the number of widows is much lower than in the other statistical areas because in this region the effects of the clashes were less devastating that in the other areas. About 10% of the total adult sampled population (almost the same percentage is found in all the four areas, see tab. 4.2.2) is disabled (12.5% of males and 9% of females). The distribution among the different kinds of disability is shown in these results: 2.8% suffer from severe vision impairments (3% of males and 2.5% of females), 0.8% from hearing/speaking impairments (0.9% and 0.7% in the two sexes), 4.2% from physical limitations (5.3% of males and 3% of females), 0.6% from mental illness (no significant difference between the two sexes) and 2.3% from not specified impairments (2.7% of males and 2.1% of females). The percentage of the adults suffering from chronic diseases is higher (17% of people living in NI: 13.6% of males and 20.1% of females). 1.4% has diabetes, 5.8% high blood pressure (with significant differences between the two sexes: 3.8% of men and 7.8% 32 Settlement and household survey report of women), 0.9% obstructive airway diseases, 1.6% heart diseases and 7.2% other not specified illnesses. Disabilities affect about 4% of the population aged under 16 (see tab. 4.2.4). Physical limitations are the most diffuse (1%), followed by severe vision impairments, hearing/speaking impairments and mental illnesses (each of them affect the 0.5% of the youngest classes of age). Chronic diseases affect the same percentage of young population (see tab. 4.2.5). In the analysis of the demographic structure of the household the figure of the household head holds particular importance since it is the person that makes decisions and that manages the income and expenditure of the whole household. Most of the sampled households are headed by men (89.5% against 10.5%). The average age is of 45 years for men and of 53 years for women, with not significant differences in the four areas. The distribution by marital status (see table 4.2.6) shows a strong preponderance of married people among the males and of widowed people among the females: 2.4% of the households heads are singles (2.4% of males and 2.1% of females), 86.5% are married (96.2% of males and 4.1% of females), 10.3% are widowed (1.1% of males and 89% of females), 0.2% are divorced (0.1% of men and 1.2% of women) and 0.6% are separated (0.2% and 3.7% in the two sexes, respectively). Finally it is interesting to notice that the great majority of marriages are based on a monogamist relationship. Even though a specific question was not brought in on this matter in the questionnaire for understandable reasons, it is estimated that only 2% of families are based on polygamy. 4.3 IDP families The issue of IDP (internally displaced people) has been extensively treated by Habitat in the IDP report7. Here we would like to report some general figures in order to give and idea of the dimension of phenomenon of people displacement in this area. The IDP report has given some estimate on the population that after displacement still live in sites and settlements that can be easily recognized as IDPs (805,000 individuals or 22.9% of all inhabitants of NI). To this population we should add, more for sake of 7 See also the IDP Site and Family Report, UNCHS, Erbil, January 2001 33 Settlement and household survey report investigation than of policy concern, those that were displaced, but that now reside in places that cannot be recognized as IDP settlements. Some of these have also return back to their homes and as such can by no means be considered IDPs. The next three tables (4.3.1, 4.3.2 and 4.3.3) give information about the answers given by household‟s heads when interviewed about their being displaced. The greater majority (2137 on 3200 hh or 66%) said that they have been displaced at least once in their life. Mostly the displacement occurred within sub district and district boundaries (see table 4.3.1). Displacements took place because of security reasons (95%). Finally, in the last table, one can see that whatever past these people experienced, for the future most of them may be considered settled since only 25% of total population are still thinking to get back to their places of origin or to some other place (district or governorate center). Only around 24% of displaced household are now living in their place of origin. This results are consistent with the outcome of the IDP study where 45% of IDPs declared themselves willing to go back to the place of origin, so to say that a certain demand for relocation and redevelopment still exists. 34 Settlement and household survey report Chapter 5 – The literacy skills An analysis carried out to study the literacy skills of the members of the household reveals the following results: - A high percentage of the NI population is illiterate (cannot read nor write) - A great number of young people never attended school - A high number of students withdraw before completing primary school - Female school attendance is much lower than male The data of the survey show that only 59.2% of the interviewed adults can read and only 58.1% can write. A strong correlation is found between the literacy skills and the age of the individual: the number of illiterate people increases, as the class of age gets older. In north Iraqi regions only 60% of adult population is able to read and write (tab. 5.1.1). Literacy is much more diffused between men (74%) than between women (44.7 %). These values are lower in Darbandikhan statistical area than in the other three. Let‟s consider the table 5.1.2. 33.8% of the sampled adult (over 15) population (20.7% of males and 46.8% of females) never attended school, 23.9% withdrew before completing primary school (24.9% of males and 22.9% of females), 19.8% (25.5% of men and 14.2% of women) finished the primary school, 11% the intermediate (14.2% of men and 7.8% of women), 5.2% the high school (7.1% and 3.3% of the sexes), 4% the institution (4.2% and 3.7%), 2.2% got a degree (3.2% and 1.3%) and 0.1% a higher qualification. Differences between the four areas are almost negligible (with the exception of Darbandikhan Area where the percentage of people that never attended school is higher), thus they may say that the problem of improving education is equally shared between all the areas of NI. At the moment 10.5% of adults are still attending school, most of which regularly whereas one third of them, as we have already said, never attended school (see tab. 5.1.3). The lack of schools in the settlements where they used to live is the primary reason that didn‟t allow them to go to school whereas the percentage of women that never attended school, more than double compared of that of men in all the four areas, can be explained with the traditional customs that expect women be solely dedicated to do domestic jobs in the home and excluded from other activities open to men. Almost 33% of women that never attended school declared that the parents didn‟t allow them to go to school (against about 10% of men) and about 35 Settlement and household survey report 15% that had to work at home (against the 10% of men). Less frequent reasons for never attending school are a lack of opportunities, an absolute refusal to go to school, the lack of importance given to attending classes, and a constant absence of teachers in the school. Although the school attendance has increased with time, today the number of the children (6-15) that never attend school still remains considerably high: 19% (tab. 5.1.5). As it is expected with relation to sex, non attendance is particularly high for females in all four statistical areas. Only in Sulaimaniyah the distance between the percentage for male and female is lower (about 7%, whereas in all other areas is approximately 14%). The distribution of sampled population in school age, according to the level of attended school (see tab. 5.1.6), indicates that 67.4% of inhabitants are attending primary school, 11.5% intermediate school, 14.3% preparatory school, 1.4% an institution and 1.9% university and 3.5% school for adults. The extremely high concentration of primary school population tells us that this is due mainly to the fact that in primary schools there are many children of higher age attending this level. Thus the schooling system is not only affected by non-attendance, but also by a great proportion of children repeating classes and thus getting inappropriate education. 36 Settlement and household survey report Chapter 6 – The wealth of households As it is widely known North Iraq population is mostly very poor. To this statement we can add that several times population rely on income sources that are not visible, particularly through a questionnaire that is more quantitative that qualitative. In particular employment figures cannot be considered exhaustive from the point of view of understanding the livelihood of local population, at least in those cases where family income is very low and the sources mostly hidden. For this reason the economic wealth of households has been detected in relation to three factors on which the survey has concentrated: employment conditions of inhabitants composing the households, durable goods and assets available to the households and expenditure patterns. The survey has produced remarkable information about these three topics that can allow us to stratify the population according to its level of economic wealth. 6.1 The employment structure It is important to understand that the employment data of the survey refer mostly to the principal occupation of the members of the households. In reality a considerable number of people have, for necessity or habit, a second job performed mostly in informal sectors of local economy. Those constitute the most evident case that besides their main occupation, for some months in the year, cultivate the land. These people generally live in urban settlements where carry out their jobs and professions and in the summer move to rural areas where they cultivate a piece of land. The household survey does not enable us to estimate precisely the category of „part time‟ farmers: the data on the occupational conditions of the household members take into account the individuals that are farmers as first occupation, so this figure underestimates the actual number of those that dedicate themselves to agriculture8. The data of the survey referred to the distribution of the sampled adult population (see tab. 6.1.1) indicates that in NI 10% of the people (13% of males and 7% of females) are employed in the Public or Government administration, 10.6% (21% of males and 0.4% of females) are self-employed, 4% (7.1% of males and 1% of females) cultivate the land as farmers or agricultural laborers, 0.4% 8 see also the studies produced by Paolo Santacroce for Habitat on rural areas in NI. 37 Settlement and household survey report (0.8% of males and zero women) are party officers, 3% are peshmerga (6.1% of males and zero women), 3.3% (6.1% of males and 0.1% of females) are manual laborers, 0.1% are artisans and 0.5% fishermen or hunters. The remaining adult population belongs to not working categories: 38.7% (76.8% of females) are housewives, 10.2% (12.8% of males and 7.6% of females) are students, 12.1% (18.2% of males and 6.2% of females) are unemployed, and 1.8% (3.2% of males and 0.4% of females) are retired. Remarkable differences can be seen between the working conditions of men and women. The active population (we can consider as active population people aged 16 and over, who have or are looking for a job) is formed by almost 85% of adult males and 15% of adult females. We have already had the chance to explain that, according to the local culture and traditions, the role of women is well defined and circumscribed to some specified activities. In a typical household the man works, earns money and maintains the family whereas the woman stays at home, carries out domestic jobs and brings up the children. So the low number of women that work or are seeking a job is explained; after few years of school a girl is set to become a housewife. 76.8% of all the interviewed women declared to be housewives and only 6.2% to be jobless (against 18.2% of men). What is curious are the significant differences that are found at a governorate level. In the governorate of Duhok 85.7% of interviewed women declared to be housewives and just 0.2% jobless, in the governorate of Erbil 82.5% declared to be housewives and 1.2% jobless, in the governorate of Sulaimaniyah respectively 64.9% and 14.3% and in the Darbandikhan Area 75.5% and 10.8%. In the governorate of Sulaimaniyah and in the Darbandikhan Area, the labor market is more open to women, probably less tied to the traditions, and the women can enjoy opportunities that are to a certain extent precluded to the women of the other two statistical areas. Whereas in Duhok and Erbil almost all the women that don‟t work consider themselves housewives, in Sulamaniyah and Darbandikhan a considerable part of the women that don‟t work consider themselves jobless. As far as the type of occupation is concerned 70.5% of the sample has a permanent occupation (tab. 6.1.2), 23.5% a temporary one and 6% an occasional one. In all the four areas the percentage of permanently working female is much higher than male (86.4% against 68.4% in the whole NI) and sharp differences are found among the regions. In the governorate of Duhok 72.9% of workers have a permanent occupation (71% of males and 90% of females), 38 Settlement and household survey report 22.3% (24% and 7% in the two sexes) a temporary occupation and 4.9% (5.1% and 3%) an occasional one; in the governorate of Erbil the permanently employed are 75.6% of the working people (72.9% of males and 93.4% of females), the temporarily employed are 17.3% (19.3% and 4.4%) and the occasionally employed 7.1% (7.8% and 2.2% in the two sexes); in the governorate of Sulaimaniyah 68.8% have a permanent occupation (66.5% of males and 83.1% of females), 26.2% a temporary occupation (28.1% and 14.3%) and 5% an occasional occupation (5.4% and 2.6%), in the Darbandikhan Area the people permanently employed represent 53.9% of the sampled working population (53.2% of men and 60.6% of women), the temporarily employed represent the 38.4% (39.4% of men and 28.7% of women) and the occasionally employed represent the 7.7% (7.4% and 10.6% in the two sexes respectively). Women are much more satisfied than men with their present occupation (55.7% against 40.6% in the whole NI - tab. 6.1.3) whereas the main cause of complaint is the scarcity of earned money, expressed by 43.9% of the interviewed workers (45.7% of males and 31.5% of females). 8.8% (8.6% of men and 10% of women) declared not to be satisfied because their job is not suitable for them, 4.6% (4.8% of males and 2.7% of females) because of other unspecified reasons and very few people were not able to express a judgment. Unemployed represent 12.1% of the sampled population (18.2% of males against 6.2% of females) with no significant differences among the four areas. 36% of jobless (43.4% of males and 14.1% of females – tab. 6.1.4) attribute their condition to the fact not to find any job, 24.6% (21.3% of males and 34.6% of females) to the fact not to find a suitable job, 22.2% (26% of men and 10.7% of women) can not work, 9.6% (2% of men and 32.4% of women) have never looked for a job, 4% (4.9% male and 1.4% female) can not find a enough rewarding job, and 3.4% (2.3% male and 6.8% female) have just lost their job. The only economic assistance to the jobless people is given by help from relatives and friends (see tab. 6.1.5). 77.1% of interviewed adults without job are receiving no assistance at all, 21.8% receive the assistance from relatives and friends, 1% from Government and 0.3% from international organizations. The retired people represent 1.8% of the sampled adult population (3.2% of the male and 0.4% of the female population). Only former government employees receive pension: self-employed and other types of workers do not receive any form of pension. For people living in North of Iraq there are two classes of retirement: the 39 Settlement and household survey report first class is constituted by individuals retired before 1991 that receive pension every three months by Iraqi government whereas the second class is constituted by those retired after 1991 that receive pension every two month by local authorities. Most of retired people (93.4% - see tab. 6.1.6) receive a pension even though in nine cases out of ten the amount of the pension does not exceed 250 ID per month and the supply is often irregular: 34.1% of retired people declared to receive pension regularly, 22.6% with rare delays, 26.6% with frequent delays not longer than three months and 16.5% with frequent delays longer than three months (table 6.1.7). The working children represent about 7% of the interviewed people aged 0-15 (see tab. 6.1.8): 5.3% (1.2% of boys and 9.4% of girls) work at home, 1.3% (1.7% of boys and 0.8% of girls) in family business, 0.6% (practically all boys belonging to very poor families) for someone outside the household. 6.2 The livelihood of the household The occupational conditions of individuals certainly constitute an important aspect of the economic wealth of the household. Besides the earnings from work there are other types of income and gains that all together form the expenditure capacity of the household. Ownership of personal and real estate (principally the property of the dwelling), loans, saved money and financial assistance are some of these types. In table 6.2.1 households have been divided by number of working members. Almost 14% of households have declared of having not even one member employed, while another 56% belong to the category of one working person per household. Interesting to notice is that the first category is much more concentrated in the area of Erbil. This people rely either on pensions or other activities that cannot be declared. The next table (6.2.2) is showing the same situation related to the family size. The first two categories (nobody employed and one person employed) are obviously more concentrated in families with less than 4 individuals. If the survey has detached 4190 working individuals in 3200 families, which makes an overall ratio of 1.3 working person per family, the same ratio changes much more in value if applied to family classes defined by size. On one side we find a consistent number of large (greater than 7) 40 Settlement and household survey report families (154) with nobody employed, while on the contrary the number of families belonging to the same size category that have 2 or more occupied members is almost the double (298). If we look at the next table (6.2.3) we can have some additional insight in the matter of income structure. In this table the monthly expenditure was weighted with the number of members of the households. Expenditure figures can be considered a good approximation of the household wealth. If we consider 100 OID per person as a significant threshold for distinguishing the population below the level of subsistence, we can see that 12.5% of households belong to this category. To a more attentive reader this may raise the question whether 100 OID may be considered the poverty line for NI. As some may remember the estimates produced by the Word Bank for the area of Middle East are much higher since they approximate the value of 1 USD of the so-called PPP (Purchasing Power Parity)9. But since in our case the PPP index is very unlikely to be estimated because of lack of information, we prefer to assume the minimum pension level granted by Local Authorities in NI as a good approximation of the poverty line. The reason why in this area the poverty line can be much lower than in other areas of Middle East (estimated PPP) is mostly due to the fact that in our case food rations are provided almost freely to local population so that the purchasing power of local population is relatively higher. It is easy to see that lower classes (below 200 OID) are much more „populated‟ in the two southern areas of the region. The suggestion of being 100 OID a rough but not negligible approximation to the poverty line is largely confirmed by the next table (6.2.4) were only those households have been selected that receive assistance (15.3% of total sample). Needless to say that relatives and religious institutions are supporting the major burden of this assistance. UN agencies and Local Authorities are much less active with respect to this matter. As compared with poverty indicators estimated in previous table assistance seems to be greater in Duhok, thanks to the fact that in this area there are more beneficiaries of remittances granted from relatives abroad. Next tables provide us with additional information about the livelihood of NI households. Table 6.2.5 reports on the distribution of households divided by the amount of money they borrowed last year. 9. See Word Development Report 2000/2001, World Bank, Oxford University Press, available at http://www.worldbank.org/poverty/wdrpoverty/report/ 41 Settlement and household survey report Interesting to notice the outstanding situation of Duhok area where we can see a lower incidence of families that have not borrow money and at the same time a higher proportion of those that belong to a higher class (more than 10000 OID). From the point of view of assets and durable goods (table 6.2.6 and table 6.2.7) this situation is confirmed. Within the sphere of durable goods 61.4% of families did declare to own none of the goods listed in the questionnaires. Just to mention some of the most important items, the car ownership is much more concentrated in Duhok and Erbil, more than half of the motorcycles are concentrated in Erbil, while animal means of transportation and agricultural land are highly concentrated in Darbandikhan. As far as durable goods are concerned, from table 6.2.7 we can infer the level of „modernization‟ of NI households. There is still some 18% of population that does not have TV sets, 45% without refrigerators and only 10% with the telephone. Looking at these figures one would infer that the availability of these modern goods decreases from North to South, placing the area of Duhok at the top of the ladder of the consumption of „modern‟ goods. If we try to draw some conclusions from the information that was gathered by the survey we may say that poverty is a very defused phenomenon in NI. If we assume the threshold of 100 OID of monthly per capita expenditure as a lowest possible indicator for poverty in the three governorates, we may see that 12.5% of population stands below the poverty line. If applied to the overall number of families (654,000) this percentage yields 82.000 families. As previously stated since this threshold is probably underestimated, it is more likely that the number of poor families in NI is greater. Anyhow this figure is in line with previous studies produced by Habitat10. Household wealth and expenditure capacity is distributed unevenly in the four statistical areas. Other indicators that were used are telling us that there is a strong regional differential between northern areas as compared to southern. The North/South divide can be seen not only in wealth indicators (income, expenditure, durable goods, assets) but also in levels of literacy, school attendance, and employment. 10 See also the IDP Site and Family Report, UNCHS, Erbil, January 2001. 42 Settlement and household survey report Chapter 7 – Housing conditions In order to analyze the housing conditions of the NI households we are going to take into account the features that synthetically represent best the adequacy of the dwelling. Housing adequacy is a concept extremely wide, multidimensional, and it varies not only according to, say, objective, measurable and technical features of a dwelling, but also to the way in which local people consider and evaluate their housing conditions. In our case we have conceived housing adequacy as the opposite concept of „housing risk‟, that is a risk to which households are exposed because of poor housing conditions. The first factor that discriminates between adequate and inadequate housing is the functional type of the building. Buildings that were not built for residential purposes cannot be considered adequate if not restructured for residential purposes. The second factor in explaining housing risk is the construction quality of the building (risk of collapsing and exposure to climate hazards). This concept can be implemented taking into account variables such as the building age and the needs of repairs. A third factor is link to the hygienic safety of the dwelling: measurable items are the access to basic services (water, electricity, sewage and sanitation system), the spatial separation of basic dwelling functions (i.e. kitchen from sleeping room, etc.) and the per capita living space. Furthermore dwellings are not adequate when two ore more families are compelled to share the same housing unit according to the principle that each family should be provided with a decent accommodation. Besides these basic components of housing adequacy concept, we can also consider unsafe and risky environment or neighborhood, uncertainty of tenure, and finally the self-evaluation of the household. 7.1 Buildings As the table 7.1.1 shows, 76.6% of the households (78.8% of those of Duhok, 68.1% of those of Erbil, 83.4% of those of Sulaimaniyah and 80.3% of those of Darbandikhan) live in onefamily unit attached to one or more houses, 12.9% (19.1% of those of Duhok, 12.4% of those of Erbil, 8.7% of those of Sulaimaniyah and 16.6% of those of Darbandikhan) live in one-family detached unit, 10.3% (1.8% of those of Duhok, 19.6% of those of Erbil, 7.8% of those of Sulaimaniyah and 2.8% of those of Darbandikhan) live in 43 Settlement and household survey report buildings with two or more housing units (apartment buildings). The remaining 0.2% lives in commercial or industrial buildings or military barracks converted in residential houses11. If we look at the building age variable we can infer that in this area the great majority of buildings are recently built. They were constructed mostly after 1970 (table 7.1.2). Half of the sampled households live in buildings built between 1970 and 1990, 28% in more recent buildings, 8% in buildings built between 1950 and 1970, 3% in buildings built in the first half of 1900 and only 0.5% in buildings built before 1900. The remaining percentage of the interviewed does not know the age of the building they inhabit. The data show some differences among the four areas. 17.6% of households of the governorate of Duhok, 23.7% of Erbil, 37.4% of of Sulaimaniyah and 34.5% of Darbandikhan area live in most recent buildings whereas 59.4% of the households of Duhok, 51.9% of the households of Erbil, 43.6% of those of Sulaimaniyah and 49.1% of those of Durbandikhan live in buildings built between 1970 and 1990. Buildings are mostly constructed in concrete blocks (tab. 7.1.3), but there is still a certain portion of housing stock built in stones (16.7%), mud or traditional and other traditional building techniques (8.3%) and others building materials. The greater majority of the buildings was built by the households that are now occupying them (7.1.4). As far as the availability of basic services is concerned the survey reveals (see table 7.1.5) that 11.9% of the buildings are not supplied with water, 33.6% are regularly supplied and 54.5% are supplied irregularly; 5.8% are not provided with electricity supply, 25% are supplied regularly and 69.2% are supplied irregularly (see table 7.1.6). Among the buildings supplied with electricity 20.2% receive electricity from public network, 35.2% from public generator, 2.3% from private generator and 42.2% alternating the public and the private system. Only 30% of the sampled buildings are connected to the public sewage system (see tab. 7.1.8); the remaining use single pit (22.4%), double pit (3.4%) and open field (74.2%) to eliminate the drain waste (see tab 7.1.9). Crossing the supply of water, electricity and sewage system with the building‟s age (tab. 7.1.5 to 7.1.9) it is found that the buildings dated back to the period 1950 – 1970 are those that are 11 See also the IDP Site and Family Report, UNCHS, Erbil, January 2001. 44 Settlement and household survey report provided with the higher number of services whereas the oldest and most recent buildings are those less supplied with the same level of services. In fact most of the houses built between 1950 and 1970 are located in the core of urban settlements and benefit more than the other buildings of those structures network present in every inhabited place of big dimension. The oldest houses were built when the great part of the water, electricity and sewage supply network didn‟t exist whereas the most recent buildings are located in outskirts or rural areas only exceptionally provided with public basic services. Besides the basic services there are other objective characteristics that improve the housing conditions. 12% of the buildings are connected with the telephone line (5.7% of those built after 1990, 14.42% of those built between 1970 and 1990, 18.3% of those built in the years between 1950 and 1970, 21.6% of those built between 1900 and 1950, 10.2% of those built before 1900 and 10.2% of those whose building age is not known (see tab. 7.1.10) even though in 18.6% of the cases the telephone line is out of order (tab. 7.1.11). The street cleaning service serves about 20% of the sampled buildings whereas the garbage collection service 65%; even in this case the buildings built between 1950 and 1970 are the most benefited (25.7% have the first service and 86.6% the second one). 7.2 Dwellings A second type of analysis that must be curried on refers to the quality and size of dwellings. The quality factor is detected by looking at the availability of basic services, at the functional organization of dwelling space and at the level of crowdedness. Availability of water within the dwelling can be considered one of these indicators. One can see (tab. 7.2.1) that there is a consistent amount of dwellings that are not provided with water inside the dwelling (16.2% of total sampled population). Consistent differences can be observed between the four statistical areas; as usual Duhok and Erbil are better off. Also interesting to mention the fact that a great proportion of dwellings supplied but not regularly are located in the governorate of Erbil. Next table (tab. 7.2.2) refers to the presence of toilets inside the building. The percentage of houses that are lacking in this facility is rather low (6.6%) and with very small differences between the four 45 Settlement and household survey report areas. This may suggest that this type of facility could be considered as a good indicator of extremely bad house quality. In our efforts of trying to discriminate adequate from inadequate housing we may take into consideration also the availability of bathroom and kitchen as an independent functional unit in the layout of the house. Tables from 7.2.3 to 7.2.5 describe the presence and distribution of bathroom and kitchen. On average 15.7% of housing units are not provided with bathroom. The dwellings located in buildings built before 1900 are those in which the presence of bathroom is scarcer. With respect to this there are two other comments worthwhile of noticing: (i) in the latest ‟90 the standard has fallen down with respect to the period that goes from 1950 to 1989, (ii) this fall has been greater in the areas of Sulaimaniyah and Darbandikhan. If we look at the data related to the presence of kitchen the situation is not far away from the one just analyzed even though the proportions my be different. The kitchen as an independent room is present in the 62.6% of the observed dwellings whereas in the remaining 37.4% is part of another room, or located in a space outside the house or shared with other households. Also in this case we may found a certain fall in the standard in recent years, particularly in Sulaimaniyah governorate. Finally some considerations about the size of the observed dwellings. Excluding the kitchen and the toilet, 9.7% of the dwellings are composed by only one room, 70.9% by two or three rooms and 19.4% by more than three rooms (see tab. 7.2.6). The dwellings built in the period included from 1950 to 1970 are generally larger. In 51.8% of the sampled buildings there is no more than 1 sleeping room, in 45.3% the sleeping rooms are two or three and in the remaining 3% of cases there are more than three sleeping rooms. In recent years dwellings have decreased by size. This may be due to the general fall in housing standards and also to a certain incidence nowadays of families with one or two members. 7.3 Housing conditions We may understand housing conditions by combining the physical and the functional profile of the dwellings with the demographic and social conditions of the household. Concepts that can help us in this analysis are “level of crowdedness”, “appropriateness of housing unit” and “family social profile”. 46 Settlement and household survey report Normally housing conditions are described by a great number of variables and indicators. At this stage we may concentrate on the ones most important. In the NI region as it is largely known people are overcrowded in their homes. If we look at the ratio of per capita available space (tab. 7.3.1) we can see that 41% of households live below the minimum standard of 10 sqm. This information may be biased by a rather rough estimate of the useful living space that surveyors were obliged to apply, since they were not able to take precise measures, but never the less this table points out a situation of very generalized housing stress. Interesting to notice that in Duhok area the ratio of households with 10 sqm or less of per capita living space is higher (52% in the governorate of Duhok, 40% in that of Erbil, 33.8% in that of Sulaimaniyah and 43% in the Darbandikhan Area) than one would expect, because of the general situation and better living conditions that characterize this governorate. One of the major reasons of overcrowdings are linked to forced cohabitation. Young couples are forced to live with the original families because they are not able to get another accommodation either by buying or by renting a new house. More than 26% of interviewed families share the same dwelling with other families (tab. 7.3.2 and 7.3.3). Also in this case Dukok stands out because of its high concentration, that can be explained neither on the bases of smaller houses nor on available income, but mainly because of stronger family relations and affection to traditional way of life. Finally some comments on ownership of land and dwellings. The survey data indicate (tab. 7.3.4) that mostly households own their house (73.6%). In recent years this phenomenon has increased with no significant differences between the four areas. In the mean time also freehold has increased while rented houses seem to account for a small proportion of the interviewed households. If we look at land ownership (table 7.3.5) and tenure (table 7.3.6) features we can see that private ownership is the most diffused type of tenure, while local authorities would mainly hold (with the exception of the governorate of Duhok) a very small amount of developed land. Freehold type tenure is very high. 7.4 Adequacy of housing At the beginning of this chapter the concept of adequacy of housing was explained and most of the tables and descriptions that followed were oriented to approach this concept by reviewing all 47 Settlement and household survey report those phenomena that are relevant to the matter and that can be understood by looking to the data of the survey. In this paragraph we will try to produce synthetic indicator derived from the data collected by the survey. The method that has been used relates mainly on the fact that adequacy of housing may be described by two groups of factors: first the quality of building and dwelling, second the way in which the household inhabits this dwelling. In order to make the analysis more transparent and understandable we have, first, developed an indicator related only to the first factor (quality of dwelling) that allows us to rank the dwellings by their quality and, second, cross-tabulated the classes obtained with information related to location, family‟s size, social level and expectations. The housing adequacy index has been obtained by taking into consideration the following attributes of the building-dwelling: type of building (as stated before non residential are considered „absolutely‟ inadequate), quality of sewage, availability in the dwelling of water, electricity and toilet, location of kitchen facility (if located in the living room or shared with other families) and finally the quality of joinery. Needless to say, this basket of variables cannot be considered exhaustive in order to fully describe the concept, but it certainly constitutes a good approximation to it. On the other hand one would need a more technical survey to get more accurate data about the layout and the quality of the houses that have been visited that in our case was out of discussion. To all of the variables described above was given a value ranging fro 0 to 1, except in the cases of kitchen and sewage that rank to a maximum value of 0.5 (in the case of kitchen because of minor importance of this factor, in the case of sewage because of lower accuracy of data). The estimated index is the sum of single scores and it may range from 0 to 5 or above in the case the dwelling is located in a non-residential building. In the first table dedicated to this phenomenon (7.4.1) we can have a look to the first outcome of this procedure. The frequency distribution approaches „normal‟ distribution. There are few values either in the low score range or in the high score range. The modal score value (1.5 with a relative frequency of 22%) is below the mean of 2.5 and important to notice, frequencies above score 3 (that is dwellings lacking in at least three items mentioned above) are extremely low. The surprising result of this table is that from the point of view of adequacy index the situation is better in southern areas (particularly Sulaimaniyah) than in the governorates of Erbil 48 Settlement and household survey report and Dukok. The reason of this unexpected distribution of low quality housing is due to historical reasons. In next table (7.4.2) the adequacy index has been grouped in 3 classes (1: smaller or equal 1, 2: between 1 and 2, 3: grater than 2). For the purpose of this study the third class is for us of greater concern. Thus in order to better understand what kind of housing profiles fall into this class a crosschecking has been made on the more frequent profiles (larger than 20) that can be found. In table H one can find graphically represented the factors mention above (SW= sewage, E=electricity, K= kitchen, T= toilet, J= joinery, W=water). Grey tones as opposed to black correspond to the actual weight of the factor within the index: Table H SW E K T J W FREQ 261 83 46 35 25 21 The quality standard of dwellings built before 1990 is generally better in southern areas, where during the „90s the same quality has sharply declined. This decline is also visible in Duhok and Erbil regions. This situation needs some comments. The survey could not account for those houses that have been destroyed (mainly before 1990) for reasons of sociopolitical conflict. Since the stock that was destroyed does not appear in our statistics, it is very likely that before that period the quality of houses in Sulaimaniyah was not better than in other areas, also because destructions occured mainly in territories with low living standards (rural areas). Thus the „substitution‟ of this stock with newly built houses has improved the overall rating in Sulaimaniyah and Darbandikhan. In this way we can explain an apparent contradiction of why we may find higher quality houses in areas where lower scores are expected. From next tables we can gather some more knowledge about the distribution of housing stock according to its standard and other 49 Settlement and household survey report variables that are linked to family and household profile. In table 7.4.3 we can see that the worse standards are present in the typology „one family house attached to one or more houses‟: this reminds the poor neighborhoods mostly located in urban areas and collective towns. An exception is Erbil where bad conditions are also found in the category „one-family house detached from any other house‟. If we check the inadequacy index against the type of tenure we can see that only slightly better conditions are found in privately owned stock, whereas rented houses and those in freehold are bad only in Erbil and Duhok governorates (table 7.4.4). As far as the size of dwelling is concerned we may find a high inverse correlation between size and adequacy index: small houses are highly correlated with worse housing conditions (table 7.4.5). On the other side it is interesting to notice that no real correlation may be found between housing quality and household size (table 7.4.6). In next two tables we tried to compare the quality of the building with the income level of the household. For income level two indexes were chosen: number of family members employed (table 7.4.7) and monthly per capita expenditure (table 7.4.8). Surprisingly we can see that relation with housing income measured in terms of employment is almost inexistent, whereas if measured in terms of per capita expenditure a strong correlation is found. Expenditure patterns are usually considered as the most reliable proxy of household income, while employment figures may not tell anything about the wealth that is coming to families from secondary activities that mostly belong to the informal sector. The worst condition is found in the upper right corner of the table. 168 households (that is 5% of the sample) belong to the category “inadequacy index greater than 2 and less than 100 OID of monthly expenditure”. To this portion of households we can certainly add those that are placed in the same column but in the next lower row that differ from the previous ones only by a higher (but still very low) index of expenditure. These account for 327 households (10%) of total population. Adding the two figures together we may get a first and very rough estimate of badly housed households that should be assisted in improving their housing conditions. The last table of this chapter (tab. 7.4.9) would largely confirm this outcome. Inadequacy index is cross tabulated against the level of crowdedness. We may find that in the worst category there are 106 households (3% of total household population) with less than 5 sqm per capita useful floor to whom we can add those immediately after, that is 316 households (10% of the sample) that belong to the same 50 Settlement and household survey report housing inadequacy category but with per capita usable floor between 5 and 10. Both tables 7.4.8 and 7.4.9 would suggest that 15% of household of North Iraq (almost 100.000 households) are in need of assistance. 7.5 Expectations After analyzing housing conditions through objective parameters we can try to understand better housing needs by looking at the self-evaluation expressed by households during the survey. The next tables (7.5.1 and 7.5.2) tell us something more about the needs and expectations. If we read carefully we can understand that housing is ranked in a very high position in the ladder of the needs expressed by the interviewed population. Looking to the first priority, at the top we find „job opportunities‟ (44%), then „water, electricity and sewage‟ (24%), after that „housing‟ (almost 21%). Summed together, these two last items will account for more than 45% of the priorities expressed, thus we can say that almost half of the interviewed consider housing and availability of basic services in the dwelling as the first priority in their needs. From these tables we can infer that besides job opportunity housing along with services linked to housing sector are far the most important needs in the view of the interviewed people. Even in the case where the job opportunity is expressed as the first priority it is usually combined as a second choice with housing or housing related items (tab. 7.5.2). From the data concerning the second priority needs we can see that housing along with basic housing services account for a great proportion of unsatisfied population, while a very important item is represented by public services. Differences between the four areas are not remarkable except for two items: households in need for job are much greater in Duhok area, whereas utilities are sought much more in Darbandikhan. In the next three tables the same data were checked against the inadequacy index applied so far. From the table 7.5.3 one can easily infer that the index is highly related to those families that look for better houses or better utilities. Only in the area of Sulaimaniyah the expectations for new houses are very low as a first priority while they increase as a second priority (see tab. 7.5.4). In the last of the tables dedicated to housing needs (see tab. 7.5.5) the same inadequacy index has been reported along with the willingness expressed by the families to participate to the process of 51 Settlement and household survey report acquiring a new accommodation. Mostly the families would participate with labor and to a lesser extent with the supervision of the building site, while a very consistent part of population are not prepared to participate to this process. 52 Settlement and household survey report Chapter 8 - Urban - Rural – Collective towns In the following chapter we will approach the problem of different social and housing conditions in the three types of settlements of our concern: urban, collective and rural. The survey data have been referred to these three categories of settlements in order to point out how different types of needs may arise in different regional context. It is evident that some of the definitions used above did not consider the urban/rural division. Neither when we tried to define the purchasing power of a family (and thus the poverty line), nor the inadequacy index of the housing unit we considered that different criteria might be applied to households living in urban, rural, or collective settlements. When one is using expenditure data to index the economic wealth of a household, he should consider the purchasing power in rural areas is relatively higher than in urban areas because of self-consumption of agricultural products. In Iraq this phenomenon might be of smaller size because of generalized supply of food rations provided by UN, but still different patters of consumptions and expenditure may give origin to different needs. Another example can be brought to our attention in relation to the inadequacy index. A kitchen that is shared between households may have a different weight in urban as opposed to rural areas where something like this can be considered also as a habit. So far we have not distinguished urban from rural and collective. In this chapter a selection of variables and indexes will be shortly presented according to their variation in these three settlements types. 8.1 Household’s profiles As everybody may know the social structure of urban (and collective) as opposed to rural population has gone through an extensive change in the last decade. Villages have been and are still resettled by a population that has experienced several displacements and if not an integration into urban society, certainly a loss of identity in rural terms. Not all of the households that resettled are now relying on farming as primary activity; some of them are shifting to more urban opportunities that they were able to take during the period of displacement. Rural areas, more than urban areas were then subject to rapid change. 53 Settlement and household survey report We may have a look to the employment figures reported in tables 8.1.1-4. Employment structure is rather different in the three settlement groups of our concern. Urban adult (over 15) population is composed of public officers, clerks, self-employed, manual laborers and, besides these, of students and retired population. In collective towns there is greater incidence peshmerga along with housewives. In rural areas besides agricultural laborers and farmers one can find a certain proportion of peshmerga and a higher incidence of housewives. Jobless population can be found in all the free settlements types consistently represented, while in collective towns the proportion of this category is higher, as it is expected. The reasons of being jobless vary from settlement type to settlement type. While in collective tows and urban centers we may find people that declare that they can not find a job or that can not work, in rural sector and particularly in southern statistical areas some will say that they have never looked for a job. If we look at the type of assistance population receive we can see that most of it is concentrated in urban areas and comes from relatives and friends (tab. 8.1.4). If we look at the table where households are divided by the number of individuals employed in a household we may see that in rural areas households with no one employed account for 17.5% of total households, whereas with one person employed account for 60.5%, in collective towns these values are 34.3% and 49.3%, while in urban areas are 24.2% and 48.2%. This means that employment in rural areas is more evenly distributed through households while in urban and particularly collective areas that are many more families either without any member employed or with more then one. In urban areas jobs are mostly of permanent or temporary type. In collective towns the temporary type increases while in rural areas both temporary and occasional may be found. 8.2 Housing conditions In all four statistical areas the recent housing stock is more concentrated in rural areas because of redevelopment that took place during the ‟90. An exception is the governorate of Dohuk were new building have also been located, because of conflicts that arise in this period near to the area of Turkish border. Housing stock of urban 54 Settlement and household survey report areas is mostly dated between 1970 and 1990, the same in collective towns (tab. 8.2.1). The types of houses are obviously also different from settlement class to settlement class. In rural areas there is a greater proportion of detached units which account for about 37% of total rural housing stock (tab. 8.2.2). With reference to the basic services supplied to housing units we may see that as far as water is concerned (tab. 8.2.3) rural areas and collective towns are in worse conditions particularly in the Sulaimaniyah governorate; electricity (tab. 8.2.4) is mostly available in urban and collective settlements while there is still a large proportion of rural houses that are without; the same consideration applies to toilets. Houses with no toilets are mostly concentrated in rural areas (tab. 8.2.5). Last three tables of this paragraph give us some insights in the question of housing size and available floor space. Houses are usually very small with no more than one sleeping room (tab. 8.2.6). As one can expect in collective towns and in the rural settlements this category is more frequent even though the shortage of living space affects all the three settlement types. While it is proportionally more relevant in collective towns it is also of great importance in absolute terms for urban areas since 36% of urban population live below the standard of 10 sqm per capita. In the next table (tab. 8.2.8) households are reported where two or more families share the same housing units. Surprisingly we may say that cohabitation is not significantly more frequent in collective towns: it is spread evenly in the three categories, with a higher concentration in some areas like Dohuk. Finally the last tables reports the distribution of housing stock in the three settlement types along with the housing inadequacy index (tab. 8.2.9). It is easy to see that most of the inadequate stock (rated more than „2‟) is located in rural areas (more than 50%). This is due mainly because of great importance given to factors like water, electricity and toilets that are in rural areas quite rare. This does not automatically mean that housing policies should mostly concentrate in rural areas providing new houses, it certainly means that some action is needed at least to alleviate the problem of some basic needs. A high concentration of bad housing is also found in collective towns, particularly in the area of Dohuk, while certain concentration it is also found in urban areas of Erbil governorate. 55 Settlement and household survey report Chapter 9 - Housing vulnerability In order to provide an assessment of the figures of vulnerable people living in Northern Iraq some additional analyses of the S&H Survey data were carried out. These analyses are better designed to depict a complex, multidimensional phenomenon, as it is the one of housing vulnerability, that needs a specific analytical approach that makes use of more adequate analytical tools. As it has been stated in previous chapters housing vulnerability is in our eyes a combination of the following three factors: - housing inadequacy or housing risk - household‟s economic wealth - household‟s demographic strength The two last factors account for what it has been defined as coping capacity. So far we have analyses all of these tree factors separately. Economic wealth and demographic structure have been considered only with respect to specific variables or indicators such as household per capita expenditure, per capita available floor, household size, dependency ratio and many others. In this chapter we will try to explore the collected data in order to get some significant household profiles that will help us in understanding how can be the concept of housing vulnerability applied to NI population. For this purpose two different research methods were applied: Factor Analysis and Hierarchical Cluster Analysis. Both of these methods belong to the family of tools of Multivariate analysis. The Factor Analysis is often used in the analysis of survey data to identify any common components or factor underlying a set of items (observations). It is helpful to reduce the variables to a smaller set of factors or to find the fewest number of variables that explain the most variance of the phenomenon. The Cluster Analysis is a technique used to identify internal structure of a phenomenon and to group together all the units of a set (a sample or an universe) that present a resemblance between each other. The input data of a Cluster Analysis can be a matrix of original values or an output of a Factor Analysis. Both of these techniques were applied to the socio-economic and demographic aspects of the households. In the following pages the socio-economic and demographic aspects were firstly studied separately, with the intent to define those factors, among the vast quantity of information provided by the survey, that, taken each one by itself or in combination with the 56 Settlement and household survey report others, were more suitable for describing the subject of our research. The selection of the first, large group of variables to be involved in the explorative analysis, the accurate examination of the amount of information brought about by each one of them into the total already provided by all the others and the definition of the last set of variables were maybe the most delicate point of the work. Secondly three classifications of the sampled households followed up: the first one according to their economic status, the second one to their demographic strength and the third one to the adequacy level of the dwelling they are living in. A brief description of the most significant characteristics of each category was provided when necessary to aid comprehension. Our attention was obviously focused on the households living at a sub standard level and this description was more accurate for the lower classes (those that present the worse conditions) and fairly rough for the other ones. Finally, the assessment of the number of vulnerable groups (as far as housing conditions are concerned) came out by crossing the outcomes of the classifications formerly produced. The following pages explain, step by step, the strategy adopted in the two multivariate analyses, summarize the results that have been produced and contain some tables that represent the different classifications and help the reader to understand the achieved results. 9.1 Economic profiles of the household In a first analysis of the economic wealth of the household many variables were taken into account, some related to the income (money earned by each worker, saved money and borrowed money during the past year, money gained from selling the dwelling) and expenditure (total monthly expenditure of the household and lend money during the past year), some to the working situation (working status of the household members and, for the workers, type of job), some to the literacy skills (school attendance and capability to read and write of all the individuals of the household). A few of them were immediately excluded from the analysis due to their unreliability: first of all the money earned by each worker, that presented too many missing values and an absolute inconsistency with the expended money, followed by the borrowed and saved money by the household and the money gained from selling the house. Then the correlation among all the kept back variables was studied, after adding a new variable (the “amount of available useful 57 Settlement and household survey report floor space in the dwelling”), thought to be a good parameter of the wealth status of the household because of its quite strong link with the amount of the monthly expenses. The weak correlation among the majority of the variables taken into account pressed us to exclude most of them from the factor analysis and treat only a) the available useful floor surface in the dwelling, b) the amount of monthly expenses and c) the number of working people in the household. In order to obtain three parameters related to the real dimension of the household, the available floor surface, the amount of total monthly expenses and the number of working people were normalized dividing each of them by the number of household members. So, the final variables (“active variables”) submitted to the multivariate analysis are: - a) the available useful floor surface per person in the dwelling, - b) the amount of monthly expenses per person - c) the percentage of working people over the households members Due to the fact that the “available useful floor surface per person in the dwelling” and the “amount of monthly expenses per person” showed the strongest correlations between each other and seemed to be, by themselves, good measurement tools of the wealth status of the household, it was thought to give a stronger weight to both of them, precisely the double to that given to the “percentage of working people over the household‟s members”. To the multivariate analysis three more variables (“supplementary variables”) were added that were not involved in the determination of the classes but that provide further information for a better comprehension of the characteristics of each class. These variables are: - the percentage of unemployed individuals in the household - the ratio between adult females and adult males in the household - the percentage of literate adults in the household Evaluation of the available data Even though the quantity of available data was satisfactory the information about the monthly expenditure was omitted from only 5 records, that were excluded from the analysis, whereas the missing 58 Settlement and household survey report values in the ratio between adult females and adult males were substituted with estimates calculated taking into account the demographic variables that showed the biggest correlation with the variable which contained the missing values. There were also 29 records of households showing very high values in the expenditure variable that were excluded from the multivariate analysis and reentered in the class 9 at the end, to avoid some negative derangements in the classification process. Multivariate Analysis outcomes The multivariate analysis has produced a classification of the sampled households in 12 classes. Here it follows a brief description of each class. Class 1 groups together the weakest household from the economic point of view and represents about one fourth of the whole sample. On average, the households have a monthly expenditure per capita that is half of the overall average of the sample and a living space in the dwelling of 7 square meters per person, but that in many cases doesn‟t exceed the 5 square meters. It is a large group of households that shows different aspects from the point of view of the demographic, past history and general living conditions profiles but characterized, at the same time, by a low percentage of workers and a low percentage of jobless people (therefore in this class there is a preponderance of households with many children and inactive people). A cross distribution by geographical area and type of settlement (see table 9.1.1) indicates that 21.7% of the households living in urban areas (25% of Erbil, 24.5% of Darbandikhan, 21.4% of Duhok and 17.3% of Sulimaniyah), 37.9% of those living in the collective towns (50.8% of Duhok, 34.4% of Erbil, 32.5% of Sulimaniyah and 26.7% of Darbandikhan), and 29% of those living in rural areas (33.8% of Duhok, 30.7% of Erbil, 30.6% of Darbandikhan and 23.1% of Sulaimanyiah) fall in this class. Class 2 groups households whose economic level is still below standard but a slightly higher than the households belonging to class 1: the average living space per person reaches the minimum acceptable level (10 square meters) but the expenditure per capita still shows poor economic condition, even though the percentage of 59 Settlement and household survey report workers is higher and the percentage of unemployed people is equal to the overall profile. The class represents 13.5% of the sample and the distribution of households, by type of settlements, that belong to this class, doesn‟t show particular differences: 13.1% of those living in urban areas, 12.5% of those living in collective towns (the top is reached by Sulaimaniyah – 17.3%) and 16% of those living in rural areas (20.2% of the households living in the rural areas of the governorate of Erbil) fall in this class. Class 3 is characterized by a percentage of workers near zero, where the adult female component predominates over the adult male one and the percentage of jobless reaches the maximum value among those of all the other classes. The average monthly expenditure pro capita is close to the lowest levels whereas the living space is just above the minimum acceptable level (10 meters). 10.4% of the sampled households fall in this class (8.6% of urban, 16.5% of collective towns and 10.1% of rural). Class 4 represents 4.3% of the sample, that is households living at a satisfactory economic wealth level, with high monthly expenditure and adequate living space in the dwelling. The percentage of jobless people and the ratio adult females – adult males are on average, whereas the percentage of working people is slightly lower and that of people that can write and read slightly higher that the overall profile. Sensible differences can be appreciated at a typological and geographical area level: 5.6% of households living in urban area (7.2% of Duhok, 5.7% of Sulaimaniyah, 5.5% of Erbil and 1.4% of Darbandikhan), 1.5% of those living in collective towns and 2.6% of those living in rural settlements fall in this class. Class 5 groups households with a sufficient economic wealth status, expressed by an individual monthly expenditure a little bit higher than the mean expenditure of the whole sample and a generally high useful living space. On the other side, the working conditions are not good at all, highlighted by a low number of employed and a high number of jobless people, and all the households components show a low degree of literacy skills. This class represents 3.5% of the sample (4% of urban households, 1.8% of households living in collective towns and 3.6% of urban households) and can be considered a cluster that place itself between economic poor households and well-to-do ones. Class 6 follows faithfully and emphasizes the general aspects of class 2. It is characterized by a sub standard economic level, 60 Settlement and household survey report expressed by the very low expenditure per capita and a just more than sufficient living space, even though the percentage of working people reaches levels touched by no one of the other economic classes and the number of adult women and jobless people is significantly lower than the overall profile. It represents 6.2% of the sampled households (6.5% of those living in urban areas, 4.8% of those living in collective towns and 6.9% of those living in urban areas). Class 7 characteristics are similar to those of class 3 (very contained expenditure level and low percentage of workers), even though the available square meters per person reach a satisfactory level. The percentage of jobless households components, the ratio adult females – adult males and the percentage of people than can read and can write reflect the overall profile. It is a quite large class (12% of the sample) that shows some differences in the distribution, either by settlement type or by geographical area. In fact it is composed by 10.8% of urban households (17% of Durbandikhan, 13.3% of Sulimaniah,, 8.8% of Erbil and 8.3% of Duhok), 13% of households living in collective towns (20.3% of Sulimaniah, 16.7% of Darbandikhan, 11% of Erbil and 5.6% of Duhok) and 15.3% of rural households (16.9% of Duhok, 15.3% of Darbandikhan, 15.1% of Sulkaimaniah and 14.1% of Erbil). Class 8 is a small class (2.2% of the sample) in which can be appreciated the characteristics of economic sound household: monthly expenditure per person over the average, exceptionally high useful living space in the house, percentage of workers almost double than the overall profile and low percentage of unemployment, a ratio adult females – adult males and a percentage of literate people at a standard level. 2.7% of households living in urban settlements, 0.8% of those living in collective towns and 2.1% of those living in rural settlements fall in this class. Class 9 represents the few, very rich households, most of them concentrated in the urban areas (3.5% over the total urban households, against 0.7% of those living in collective towns and 0.7% of those living in rural areas). The square meters per person, the monthly expenditure and the percentage of workers are all remarkably high, as well as the percentage of literate people. The unemployment rate is on average whereas the adult females – adult males ratio is very low (only in this class goes below the level of parity). 61 Settlement and household survey report Class 10 like class 5 is a cluster that ideally separates the economically sound households from the economically weak ones. As the class 5 has an over standard living space per person and an average low expenditure per capita, but at difference of that it shows a fairly good percentage of workers and literate people, and a standard unemployment rate. It represents 6.1% of all interviewed households: 7% of the urban ones, 4% of those living in collective towns and 5.5% of those living in rural areas. Class 11 is a cluster that presents some contradictory aspects. In fact the monthly expenditure per capita is greater than overall average but the square meters per person available in the dwelling are not but sufficient, the percentage of working households components is under average even though the literate rate is high and the unemployment rate and the ratio adult females – adult males are at a standard level. The large component of households living in cities or towns (11.3% of urban households fall in this class, against 5.8% of those living in collective towns and 6.5% of those living in rural area) can be an explanation for this apparent inconsistency. Even if the expenses are higher than average the living space in the dwelling remains low because of the general lack of living space present in the urban centers. Class 12 represents 3.8% of the sample and groups a category of households that enjoy a good economic wealth level as all the parameters taken into consideration testify. The adequate living space in the dwelling, the monthly expenditure double than that of the overall profile and the very high employment rate indicate that the general economic level of these households is above the standard. The low percentage of jobless people, the high literacy rate and the ratio adult females – adult males just below the average are a confirmation of that. The distribution among the types of settlements shows these figures: 5.2% of the households living in urban areas, 0.5% of those living in collective towns and 1.9% of those living in rural areas fall in this class. 62 Settlement and household survey report Table I - Description of the 12 economic wealth classes produced by the factor analysis CLASS 1 NUM weight Active variables Supplementary variables perc. 834 26.1% 2 432 13.5% 3 331 10.4% 4 137 4.3% 5 113 3.5% 6 199 6.2% 7 383 12.0% 8 71 2.2% 9 80 2.5% 10 196 6.1% 11 299 9.4% 12 120 3.8% OVERALL PROFILE 3195 2 m2 7.27 <<<< 10.15 << 200.23 << 33.37 >> 11.76 << 180.40 << 0.69 <<<< 23.40 == 877.17 >>>> 16.61 << 36.18 >>>> 325.49 == 6.20 << 13.14 == 212.52 << 56.32 >>>> 16.99 == 186.16 << 19.23 << 59.72 >>>> 471.33 >> 41.18 >>>> 31.74 >>>> 1433.60 >>>> 37.78 >> 29.02 >>>> 326.05 == 34.46 >> 12.75 == 469.67 >> 19.31 << 20.29 >> 595.29 >>>> 52.82 >>>> 15.05 2 expenditure 149.25 <<<< 288.69 1 % workers 15.71 << 22.88 % jobless 4.48 << 7.05 == 102.83 << 56.88 == 19.61 >> 139.20 >>>> 43.22 << 9.84 == 121.21 == 67.38 >> 17.30 >> 120.22 == 45.73 << 3.36 << 101.17 << 56.15 == 5.24 == 117.78 == 57.93 == 4.38 << 115.34 == 60.84 == 6.82 == 98.45 <<<< 74.79 >> 5.38 == 111.18 == 66.92 >> 7.79 == 118.24 == 64.69 >> 3.49 << 108.05 == 73.70 >> 7.48 ratio fem-mal 122.34 == 117.38 % literate 53.75 == 57.04 <<<< << == >> >>>> the class value is much lower than the overall profile the class value is lower than the overall profile the class value reflects the overall profile the class value is greater than the overall profile the class value is much greater than the overall profile 63 Settlement and household survey report 9.2 Demographic profiles of the household The second factor affecting the coping capacity of a household is the strength of its demographic structure. The demographic soundness of a family nucleus is assessed taking into consideration simultaneously the demographic characteristics of all its components. The great availability of data collected during the survey allowed us to study the following factors, believed to be enough to provide a satisfactory evaluation of the strength of the household demographic structure: - the composition by age and sex of the household members, that determinates, in a presumptive way, a general capability of the household to produce income and to successfully face hazards that could strike it (in the case under examination the housing inadequacy); - the presence in the household of disabled or chronically diseased people, that is individuals that can not provide to themselves (or that have a very limited capacity to provide to themselves) for reasons that depend on neither their age nor their sex and that weaken the resilience of the whole household; - the marital status of the adult females, that assumes significant importance in a country like Iraq, where the general female condition, firmly bound to the traditions and the costumes of the past, is still relegated to a background position, and where widows, divorced and separated women are likely to become vulnerable groups. We started analyzing a large number of variables (for each household we considered the number of members in each age group, the number of males, the number of females, the number of disabled adult males, disabled adult females and disabled children, the number of adults and children suffering from chronic diseases, the number of widows, divorced females and separated females) determining their frequencies and synthetic indexes –mean, mode and variance - and calculating the correlations between each other. In order to reduce the number of variables to submit to the factor analysis without loosing much information we determined some synthetic indexes that could well express the demographic structure of a household. The indexes (each one expressed either as a ratio between two household members groups or as a percentage of the total number of household components, applying the same standardization method we used in the analysis of the economic 64 Settlement and household survey report wealth status of the household) used as “active variables” in the multivariate analysis were: - percentage of household members aged 0 - 5 - percentage of household members aged 6 - 15 - percentage of household members aged 16 - 40 - percentage of household members aged 41 - 65 - percentage of household members aged more than 65 - ratio between members aged 0 – 15 and members aged 16 – 65 (dependency index for young people) - ratio between members aged more than 65 and member aged 16 – 65 (dependency index for old people) - ratio between females aged 16 – 65 and males aged 16 – 65 - average age of children living in the household - percentage of people suffering from disabilities or chronic diseases (giving weight 1 to weak males and ½ to weak females and weak children) - percentage of widows, divorced and separated females The following few other indexes were inserted in the analysis as “supplementary variables” to understand better the final classification: - ratio between members aged 0 –15 and male members aged 16 – 65 - ratio between members aged more than 65 and males aged 16 – 65 - percentage of adult males suffering from disabil. or chronic diseases - percentage of adult females suffering from disabil. or chronic diseases - percentage of children suffering from disabilities or chronic diseases Evaluation of the available data The survey provided satisfactory information about the demographic characteristics of the household components and all the variables taken into consideration in the multivariate analysis presented a good level of reliability and consistency. The few errors found in the database were corrected during the checking/monitoring phase. Some problems arose later, when the synthetic indexes were computed (ratio between females and males aged 16 – 65, ratio 65 Settlement and household survey report between people aged 0 – 15 or more than 65 and male components aged 16 – 65) and the denominator was zero. Furthermore every time the denominator was zero an error occurred and a system missing value was produced. All these missing values were substituted with estimates calculated taking into account the demographic variables that correlated better with the variables presenting missing values. A prior control on this technique showed very pleasing results. Multivariate Analysis outcomes The multivariate analysis has produced a classification of the sampled households in 8 clusters. Here is a brief description of each class. Class 1 represents 28.5% of the sample and groups large middle-aged households, headed mostly by middle-aged men, with a strong demographic structure. The average household size is of 8.6 people and the distribution by age shows very low percentages of youngest (the average age of children is around 10.5 years) and oldest components and a concentration in the classes of age 6 – 15, 16 – 40 and 41 – 65. The dependency indexes are both low and the ratio active females – active males indicates that there is a slight preponderance of active men (on average, there are 9 females aged 16 – 65 for every 10 males of the same age). The percentages of disabled or chronically diseased people and widows, divorced or separated women do not show significant differences from the overall profile. This can be considered the strongest class or the one less affected by vulnerability factors. Class 2 is a cluster of youngest households where the household heads are usually young married men (90% of them do not exceed the age of 40 years) and there is a strong preponderance of children aged 0 – 5 (the average age of children is of 3 years). On average a household is formed by 5 - 6 people mostly concentrated in the classes of age comprised between 0 - 5 and 16 - 40, with a low percentage of old people. The high incidence of children aged 5 or below makes this class vulnerable. The percentage of disabled or chronic diseased members is at average level whereas the ratio independent females – independent males is low and the presence of widows, divorced or separated female is extremely low. 66 Settlement and household survey report This class represents 18.2% of the sample. Class 3 has a distribution by age clearly unbalanced toward the youngest classes of age (almost 7 members over 10 are aged between 0 and 15 years). Consequentially the dependency index for the young population is remarkably high (the dependency index calculated putting at the denominator only the active males shows that there are 5 children for 1 active male), on the contrary, that for old population remarkably low. It represents 17.9% of the sampled households, where, on average, the family nuclei are formed by 8 components, the head of the household is a married man aged 38, the ratio independent females – independent males, like the average age of children and the percentage of weak people, respects the overall profile and the percentage of widowed, divorced and separated females is lower than average. Class 4 is a sound group of households (15.9% of the sample) that shows a distribution by age concentrated in the independent population classes (68.4% in the class between 16 and 40 years and 21.4% in that between 41 and 65). Normally these households are formed by 5 adults and no more than 1 child, whose head is a man (75%) or a woman (25%) aged from 40 to 65. The high number of independent components, the ratio females – males reflecting the overall profile as well as the percentages of disabled or chronically diseased people and of widows, divorced and separated women at an average level is a proof of the demographic profile of this class and its low level of vulnerability. Class 5 represents 7.9% of the whole sample. Seven people, whose one and half person aged 65 or more, generally form the households falling in this class. It is a cluster where there is a strong component of old people (the average age of household head is 56 years and the dependency indexes indicate that the weight of elderly is strong and that of the youth lower than average) and where the children age is slightly unbalanced toward the lowest values as their average age (6 years) shows. Even though the majority of these family nuclei is headed by males, the ratio females – males shows an evident preponderance of females and the percentage of weak females indicates the presence of a quite big number of widows, divorced and separated women. Class 6 represents a small group (3.4% of the sample) of very old households. The average size of these households is slightly greater than 3 people: generally two parents (whose household head is aged 65 and over in 85% of the cases) and an adult son (the 67 Settlement and household survey report significantly low ratio independent females – independent males indicates that an adult male generally provides to the support of the family). The high elderly dependency rate, the high percentage of disabled or diseased people and the remarkable incidence of widows (20% of the households falling in this class are led by a widow) make this cluster extremely weak from the point of view of the demographic structure. Class 7 represents 6.6% of the sample and is characterized by the high component of women: on average in each household there are more than 3 females for every male. The distribution by classes of age shows an overall similarity with the average profile; only the class aged more than 65 is slightly lower than the average value of the sample. On average a household is constituted by 7 individuals headed by a man eight times over ten. Class 8 is the smallest cluster (1.5% of the sample) and groups the households in which a high percentage of widows predominates. The average household dimension is of 4 people and in 97% of the cases the head of the family is a woman aged, most of the times, between 40 and 65 years. The distribution by age indicates two relevant points: the children are concentrated in the upper age class (that is confirmed by their high average age, that is greater than 10 years) and the active people in the class of age 41 – 65 (while the percentage falling in the class 16 – 40 is remarkably low). These characteristics confirms the hypothesis that the majority of these women lost their husbands not that recently, during the internal and external conflicts of ten or more years ago and that the condition of the households they belong to are not as weak as they were time ago. Nevertheless the high dependency indexes and the high percentage of disabled or chronically diseased members show a strong weight of people that are not able to provide for themselves. 68 Settlement and household survey report Table J - Description of the 8 demographic classes produced by the factor analysis CLASSES 1 NUM perc. 912 28.5% 6.54 << % 6_15 32.14 >> % 16_40 40.01 == % 41_65 Active variables Supplementary variables 20.34 >> % ov_65 << rat. y_act 71.75 << rat. old_act << rat. f-m active avg. chldn age weak memb. weak fem. == 60.36 == rat. y_mact 135.98 << rat. old_mact << rat. f-m 94.87 << weak_m 11.26 == weak_f 11.37 == weak_c == <<<< << == >> >>>> 4.61 3.33 90.77 << 10.55 >> 19.25 == 3.10 1.78 0.97 2 583 18.2% 41.49 >>>> 7.49 << 47.69 >> 3.00 << 0.34 << 106.62 == 0.73 << 104.73 == 3.08 <<<< 18.98 == 1.40 << 50.69 << 216.57 == 1.28 << 124.34 == 10.79 == 7.66 << 8.72 == 3 574 17.9% 22.97 >> 45.23 >>>> 26.02 << 5.19 << 0.59 << 232.36 >>>> 2.13 << 116.00 == 7.71 == 26.30 == 1.28 << 31.21 <<<< 491.90 >>>> 4.64 << 117.53 == 13.20 == 12.13 == 14.09 >> 4 509 15.9% 4.12 << 3.50 << 68.42 >>>> 21.35 >> 2.61 == 10.03 <<<< 3.37 << 111.30 == 6.46 << 18.90 == 8.73 == 89.77 >>>> 21.54 <<<< 6.92 << 105.98 << 10.98 == 14.90 == 0.92 << 5 254 7.9% 14.08 == 14.62 << 40.80 == 10.17 == 20.32 >>>> 77.73 << 42.07 >>>> 143.46 >> 6.03 << 26.50 == 10.13 >> 50.98 << 180.45 == 92.44 >>>> 137.09 == 15.27 == 19.51 >> 2.94 == 6 109 3.4% 3.18 << 4.02 << 21.22 <<<< 12.39 == 59.19 >>>> 52.43 << 105.30 >>>> 82.42 << 4.89 << 38.85 >> 16.02 >> 33.61 <<<< 141.46 << 133.30 >>>> 114.45 == 22.61 >> 31.52 >> 0.97 << 7 212 6.6% 14.27 == 23.51 == 45.40 == 15.37 == 1.45 << 85.68 == 2.40 << 267.62 >>>> 7.84 == 15.79 << 8.42 == 60.77 >> 269.18 >> 7.76 << 338.93 >>>> 6.17 << 14.56 == 4.69 == 8 47 1.5% 2.99 << 44.72 >>>> 19.55 <<<< 27.36 >> 5.38 == 186.21 >> 20.25 >> 122.90 == 10.04 >> 43.79 >> 91.04 >>>> 46.91 << 299.24 >> 22.13 == 145.21 == 0.94 << 59.28 >>>> 26.42 >> 6.63 13.70 11.66 127.21 15.53 213.12 56.12 6.13 21.82 7.32 117.41 8.73 100.47 4.54 13.35 42.77 22.99 16.36 OVERALL PROFILE 3200 % 0_5 % 16_65 the class value is much lower than the overall profile the class value is lower than the overall profile the class value reflects the overall profile the class value is greater than the overall profile the class value is much greater than the overall profile 69 Settlement and household survey report 9.3 Vulnerable households So far we have analyzed separately all those factors that if grouped can give us some insights in the matter of vulnerable groups. In order to simplify the analysis that follows the classes envisaged on the bases of multivariate analysis have been grouped in a smaller number of clusters. First the economic classes have been grouped in three clusters: - Claster 1 encompasses only one class, namely class 1 that is composed of very poor and overcrowded households. These groups are the most important for policy targeting (high priority). - Cluster 2 aggregates all the classes (2, 3, 6, 7) that include poor but not overcrowded families. This cluster should be also of some concern for decision makers. - Cluster 3 covers the remaining classes that are the ones with high income (4,5, 8, 9, 10, 11, 12). It is assumed that housing policy can neglect these households. In the table 9.3.1 one can see how these clusters are distributed in the four statistical areas and in the three settlement types. The areas that are more effected by poverty (cluster 1) are collective towns and rural areas: rural areas are more affected in Darbandikhan and Duhok, while collective towns more in Duhok and Sulaimaniyah. It is important to notice that not only no area is immune from the phenomenon of poverty (as expected), but also that cluster 1 is fairly well represented in urban areas. The same reduction has been done on demographic classes. In this case the only two sound classes (class 1 and class 4) have been added to cluster 2, while all other classes (2, 3, 5, 6, 7, 8) have been assigned to cluster 1. This cluster gathers all the households with some kind of “demographic handicap” being this handicap related to any kind of abundance of dependent members (young, old, female, widows…) over independent This clustering could have been done in a more accurate way by grouping the original classes in a greater number of clusters, but a simplest solution has been preferred for sake of simplicity of the exposition. The distribution of clusters is reported in next table (table 9.3.2) from which one can see that the proportion of the two clusters is almost evenly distributed in the three settlements types with a slighter concentration of handicapped households in collective towns and rural areas. 70 Settlement and household survey report A final set of tables (tables 9.3.3, 9.3.4 and 9.3.5) has been produced by combining adequacy level of housing unit, with economic and demographic clusters. The combination of these three clustering criteria can allow us to rank the households profiles according to their level of housing vulnerability. In table L household clusters are reported not only in sample terms, but also exploded to the universe (see projected HH). The figures are impressive. Only the first vulnerability group accounts for 7.17 % of total household, that is for 46 thousands units. The first four groups, that are those of the most vulnerable, account for 130 thousands units. It is important to notice that the estimates of vulnerable households should not be taken as estimates of needed housing stock in the sense that we can not take as granted that each vulnerable household should be provided with a new house. New houses should not mechanically replace the ones that are occupied at the moment they should be used in order to house the most vulnerable and to induce in the rest of the stock a „filtering up‟ process through which every household could improve its housing condition. 71 Settlement and household survey report TAB. K: Sampled and projected household ranked by vulnerability status Sampled Cumul. Cumul HH HH % Highly inadequate, overcrowded, very poor, demographic handicap Highly inadequate, overcrowded, very poor, no handicap Inadequate, overcrowded, very poor, demographic handicap Inadequate, overcrowded, very poor, no handicap Highly inadequate, poor, demographic handicap Highly inadequate, poor, no handicap Inadequate, poor, demographic handicap Inadequate, poor, no handicap Adequate, overcrowded, very poor, demographic handicap Adequate, overcrowded, very poor, no handicap Adequate, poor, demographic handicap Adequate, poor, no handicap Highly inadequate, high income, demographic handicap Highly inadequate, high income, no handicap Inadequate, high income, demographic handicap Inadequate, high income, no handicap Adequate, high income, demographic handicap Adequate, high income, no handicap 229 96 199 117 199 163 289 249 112 81 213 235 72 68 214 179 249 231 229 325 524 641 840 1003 1292 1541 1653 1734 1947 2182 2254 2322 2536 2715 2964 3195 7.17 10.17 16.4 20.06 26.29 31.39 40.44 48.23 51.74 54.27 60.94 68.29 70.55 72.68 79.37 84.98 92.77 100 Projected HH 46588 66119 106604 130407 170892 204053 262848 313505 336291 352770 396103 443912 458560 472394 515931 552347 603005 650000 Urban Collective % % 1.94 3.16 6.7 8.64 10.39 12.27 17.21 21.75 24.73 26.79 32.18 38.28 39.09 39.87 44.66 49.23 56.31 62.94 2.57 3.54 5.41 6.51 8.33 9.89 12.08 13.96 14.27 14.52 15.27 15.9 16.31 16.68 17.59 18.25 18.65 18.81 Rural % 2.66 3.47 4.29 4.91 7.57 9.23 11.14 12.52 12.74 12.96 13.49 14.12 15.15 16.12 17.12 17.5 17.81 18.25 72 Settlement and household survey report Chapter 10 - Final recommendations NI has experienced a great transformation in last ten and more years. This area has transformed quickly from a rural society to an urban without being developed in the mean time the needed urban structure: that is those basic activities that make urban life sustainable. Thus NI has had a turn from rural society to a “semi urban” while the same rural areas have suffered a profound decline. Poverty is spread in the households of the three governorates and it becomes more and more intense when we go from Northern to Southern areas. 12% of the interviewed households have apparently no formal income source (from salary or wage) while 45% of them have declared to have a per capita expenditure of less than 200 OID per month (just above the poverty line). Some 40% of the households live in housing conditions where individuals have less than 10 sqm of per capita useful floor, while 25% of the housing stock can be considered inadequate to house a family at a required standard. These figures give us an impression of an area under profound and generalized stress. We may not know how many were the houses and activities lost because of the many conflicts that have affected this area, but we can understand that a great damage has been brought and that the area is still far away from recovering. One of indirect indicators of the distress suffered but NI population is the fact reported in the survey that 66% of the interviewed have declared that at least ones in their life they had been displaced. Another important observation is that poor conditions are not only found in IDP sites and collective towns, but there are consistent groups of needy households as well as in urban and in rural areas. If we add to this that a relevant proportion of NI population that had been displaced is still thinking how to get back to their place of origin, we can understand that emergency condition are the ones prevailing. This means that policy targeting becomes extremely important since we can not say, as it has been thought for a long time, that distress affects mostly one category of local population: namely the IDPs. Poor settlement and housing standards are found everywhere in this area. Which means that the resettlement program should not only be kept going on, but also been given a broader domain of humanitarian action. 73 Settlement and household survey report Poverty is manly due to the fact that production levels are very low. Even though the rate of unemployment cannot be considered as extremely high for a situation like this (12 % of active population), income levels are usually very low. In such a situation it is difficult to distinguish between households that live on the basis of the work performed by their members from those that live upon other financial sources that can not be acknowledged formally (aid provided by relatives from abroad or local religious institution, income gained through activities that are not declared). For this reason is also difficult to distinguish households according to their wealth level. The survey has found that several households with no income source present expenditure patterns that are those of wealthy families. The housing conditions are characterized by a great diffusion of overcrowdings and lack in basic services. In rural areas and collective towns these bad conditions are more frequently found, but also in urban areas there is a certain concentration of inadequate housing. Most of the households face the problem of better housing either because they are overcrowded or because of poor services supplied. Both these factors can be explained because of the emergency type situation. Thus it is very important to understand who are the households that are more than others able to cope with this problem. This study has pinpointed both the factors and the profiles of household vulnerability. The major factor is the lack of working opportunities, particularly in those cases where unemployment is an endemic phenomenon. This can be due to several reasons: 1) to lack of available manpower in the household (young men have left or disappeared and women are not expected to enter the labor market); 2) to lack of opportunities caused by social distress that in some cases has effected whole generations that are not even willing to look for a job anymore. Another factor of vulnerability is the demographic structure of a household. There are plenty of household types that have been envisaged by the analysis that face the problem of a very high dependency ratio. This can be due to different households‟ profiles. There are households with predominance of elderly people, or of female members, or of children in tender age, or of widows that have to take care of their children and their parents. Because of many misfortunes suffered by this population, the social organization of family relationships does not guarantee any more a balanced composition of weaker members with active population. Households are either too old, or too young, or too feminized. In all of these cases 74 Settlement and household survey report some assistance is needed in order to increase the coping capacity of local population. These are the major factors. In some cases the interviewed have complained about security, unsuitable neighborhood and other environmental factors. But these are far less relevant than the ones mentioned above. Looking at this situation that has been sketched above the resettlement project run by Habitat should clearly encompass some policy objectives of general nature, while being more precise in pursuing some specific and selected policy targets. 10.1 Policy review The policy objective of a humanitarian program cannot be only directed towards the provision of service and stock delivered to needy beneficiaries; it has to consider and match more carefully the complex structure of needs of local inhabitants along with the possible outcomes of humanitarian action. a) Building new houses and infrastructures require a more comprehensive approach The survey shows how the housing stock and the services supplied to housing (settlement infrastructure) have to increase dramatically if we want the local population being released from the present emergency situation. It is obvious that both housing and infrastructure have to move side by side if we want to achieve improvements simultaneously. In some situations a consistent intervention is needed in order to upgrade housing in terms of adding basic services and infrastructures more than building new houses. In some other cases new infrastructures can encompass the needs not only of the site they were planned to belong to, but also to a larger service area since a substantial increase in service standards is needed everywhere. Thus the future humanitarian action of resettlement programmes should relay on specific assessments (possibly produced for each area or community) so to make clear which cases may be dealt with new housing (or housing rehabilitation), which with additional service supply and which with both12. b) Program’s overall impact 12 On this matter see also the SSI evaluation report 75 Settlement and household survey report Housing demand in NI steams mostly from housing inadequacy, overcrowding and low coping capacity of inhabitants. Humanitarian action should face all this problems together. Building new houses does not mean to give relief only to selected beneficiaries, but it should bring to a general result of decreasing crowdedness in the rest of the housing stock. This can be done only under one condition that allocation policy is based on the family unit and not on the household unit. This may cause problems in NI society where the habit of „extended family‟ is still very strong and where cohabitation between families is not always forced. It can also be of some risk in an area where security of tenure is a very blurred concept or at least the population perceives it like this. Some studies have already shown how new Habitat houses are inhabited in the same overcrowding conditions as the previously occupied housing stock13. The second important aspect linked to the effects of resettlement program on population needs refers to the impact of program‟s building activities on the labor market. From our survey comes out that people ask above all for jobs and housing. Most of the families that are in need of a dwelling are also prepared to participate to the building process by labor. This type of coordinated action with final users it has been already practiced by Habitat with self built packages. What is rarely been practiced is the coordination with other agencies that can also directly or indirectly provide for income generation activities. The survey shows that there is more room for this type of intervention and what kind of family profiles can be attracted by this type of policy action. Both these aspects require additional means for project management (community development, infra and inter agency coordination, tenure system definition….) that are not available at the moment. The same concept of housing or settlement „project‟ should be reviewed in order to meet these general requirements, otherwise the humanitarian action could be largely considered ineffective14. c) beneficiaries profiles 13 14 see IDP report. Some suggestions have already been formulated in the SSI report. 76 Settlement and household survey report In this situation an accurate selection of project beneficiaries is of great meaning. The selection criteria of beneficiaries has not only to be derived from the immediate effects expected on directly assisted population, but also to the side effects on the whole settlement and housing conditions of this area. The survey has pointed out at least two aspects that are worthwhile of being mentioned. First, we have finally obtained an accurate description of vulnerable groups (households), that shows us to what extent the terminology used so far in policy implementation is inadequate and obsolete. We now know better where the poverty line can be placed and what are the factors affecting vulnerability of households. We also know something more about the possible reactions of different household profiles to policy measures and types of packages supplied. If we compare the results of this study to the current terminology used in project allocation we can see that some of the stereotypes used so far are not valid any more. For example we cannot mechanically assume some of the „traditional‟ vulnerable groups, like widows or elderly population, as the most vulnerable. In both these cases the survey has shown that vulnerability is a mach more complex term since it encompasses: - Poor or inadequate housing, - Overcrowding - Weak demographic structure - Low economic level - Low education and school attendance of children Useful indicators to depict vulnerability have shown to be: available living space per capita, the dependency ratio, the sex ratio, expenditure patterns, school attendance and other indicators. Not all of the indicators used in this analysis can be mechanically transferred to the operational field of housing allocation. From this point of view some additional work is required. But it should be clear now who are the most vulnerable in this area, who are the ones affected by a combination of economic, demographic and housing problems, and finally who are the ones that because of their poor condition do not send their kids to school. Second we now understand that vulnerable groups are not easily to be identified in a situation where low income and poor living conditions are prevailing. The muddle between vulnerable and poor household lies always in wait. Our analysis has shown that being poor is a necessary but not sufficient condition for being vulnerable. It has also shown that vulnerable population may be 77 Settlement and household survey report found everywhere: in urban areas, in collective towns and in rural territories. This means that a policy response must be tailored to these different geographical profiles and that we can not apply the same project criteria to urban or to rural population, to IDPs and not IDPs, to those that are vulnerables because of high dependency ratio of young members or of old members or because of high feminization of the household. We may guess that poor housing condition in urban areas is less sustainable than in rural areas, whereas poverty (low income) could be more sustainable in urban centers if a labor force is available within the household, it is the opposite for household composed by old individuals, unable to work and that do not receive any assistance. d) Need for resettlement schemes The way in which resettlement is in NI needs to be rethought over Habitat delivery system cannot be considered anymore as a supply of shelters and other services for people that have lost their homes (reward). It has to be viewed as one means (between many) of stopping the on going process of impoverishment of local population. This process is more likely to be explained on the basis of general economic conditions produced in recent times than with respect to old political conflicts that stroke the area. In North Iraq no family is without a roof, but there are many families that would be absolutely helpless in trying to improve their housing condition which is seldom inhuman or far beyond any imagination. The lesson that comes from the results of this survey is that problems can and should not be addressed in a isolated way, mostly by providing new housing, but in a coordinated fashion. People are not only vulnerable because of inadequate housing, but because they do not have jobs, because women have been left alone, because young families have to many children, because kids do not attend school, because there is shortage of water, electricity and all other services that can make the life of these people sustainable. All this problems have to bee addressed with projects coordinated by resettlement schemes 78 Settlement and household survey report Appendix A: S&H Survey Team 1. Igor Jogan, Survey Team Supervisor (senior consultant) 2. Monica Noro, Survey Team Coordinator (consultant) 3. Paolo Santacroce, Rural Area and Data Processing Senior Consultant 4. Rado Fonda, Statistical Consultant 5. Marco Carbonara, Junior Planning Consultant 6. Valerio Antonietti , Junior Anthropological Consultant 7. Giorgio Plazzi, Data Processing Consultant 8. Dalya Qahraman, Survey Assistant 9. Peshkawt Wali, Database Designer 10.Azhee Abdulsamad, Data Entry 11.Herish Musheer, Translator 12.Serwan Merza, Data Entry 13.Gushad Mawlood, Admin Clerk 14.Rizgar Maghdid, Database Operator 15.Muhammad Bilbas, C.M.O. Erbil Field Office 16.Yousif Othman, C.M.O. Suly Field Office 17.Ahmad Muhammad, C.M.O. Dohuk Field Office The field team was composed by 40 surveyors, 4 data entry and 4 data checkers. 79 Settlement and household survey report Appendix B TABLES 80