Abstract: A questionnaire survey was conducted among a group of factory workers in Addis Ababa to assess the extent to which public ownership of land has promoted low-income home ownership. The findings indicate that approximately one-fifth of the respondents have become homeowners mainly due to the supply of free building lots. Although reported interest in home ownership was quite high among the remaining respondents, their attempts to take advantage of the supply of free land have not been as impressive. Discriminant analysis results suggest that the decision to file applications to secure building lots is mainly a function of stages in the workers' life cycle and the level of satisfaction in the quality of their dwellings rather than of their awareness of the availability of free land.
In the recent past many developing countries have adopted national urban housing policies and programs in order to improve the housing plight of their low and moderate income households. One of the African nations that adopted profound urban housing policies during the last two decades is Ethiopia. The new housing policies that were introduced by the country's Marxist government in mid-1975 were so radical that they involved the nationalisation of all urban land and rental houses. In this paper a modest attempt is made to shed light on the extent to which this fundamental change in urban housing strategy has promoted low and moderate income home ownership in Addis Ababa, the national capital.
It is well known that millions of people in the cities of the Third World are either homeless or inadequately housed. It is also evident that this housing crisis is worsening day by day. From the public policy perspective it appears that at least two main factors have been responsible for this intractable shelter problem. In the first place, most governments did not attempt to arrest their deteriorating urban housing problems until it was too late. Indeed, many Third World countries, and especially African states, did not adopt substantial urban housing policies and programs until the early 70s. Secondly, when they eventually did, most governments introduced very unrealistic housing policies and programs (Lea, 1980; Khabagambe and Moughtin, 1983; Mabogunje, 1990; Stren, 1990; Salau, 1990). In the recent past, a growing awareness of the shortcomings and consequences of the urban housing strategies of many developing countries has led to an increased search for reliable and cost-effective ways of improving the housing plight of the working poor. As a result both liberal and neo-Marxist scholars have attempted to provide a panacea for the urban housing problems of the Third World. Liberal writers have persistently argued that encouraging the working poor to build their own homes based on the principles of self-help housing is the best strategy to increase the supply of low-cost housing (Turner, 1976; 1979; Lea, 1980; Stren, 1990). Comparatively, neo-Marxists have contended that since housing is a dependent subsystem of the broader socio-economic structure, low-income shelter needs can be meaningfully met only through the radical transformation of the socio-economic formation in question. Indeed implicit in nearly all Marxist literature on urban shelter problems is the assumption that the capitalist system has to be removed if the equitable distribution of resources, including housing, is to be realised (Burgess, 1979; Pezzoli, 1987; Nientied and van der Linden, 1990).
Aside from their significant differences over the philosophy of self-help housing, both liberal and neo-Marxist scholars have at least one fundamental area of implicit agreement in their approaches to low-income housing. This area of tacit agreement occurs in the position that either group has directly or indirectly taken with regard to the need for public control of urban land. Indeed, one might argue that the primary interest of the liberal theoreticians is to maintain the status quo. However, the fact remains that the large-scale implementation of self-help housing schemes that they advocate cannot be realised without some kind of redistribution of vital resources, including land (Devas, 1983, Pezzoli, 1987, Asiama, 1990).
During the 70s and 80s several developing countries experimented with either the liberal or the neo-Marxist strategies of building houses. However, neither the two approaches nor the combination of elements of both housing strategies has succeeded in improving the housing problem of the urban poor. This has been particularly the experience of most capitalist and socialist African nations where self-help housing programs have invariably tended to subsidise the middle and upper income households (Khabagambe and Moughtin, 1983; Macoloo, 1988; Main; 1990; Stren, 1990; Mathey; 1990). In some instances, public control of urban land has even served to aggravate the housing problems of the urban poor (Salau, 1990; Jenkins, 1990).
Some African countries have recently tried to increase urban housing supply by decentralising the decision-making process and by encouraging popular participation in the development of urban infrastructure. However, such shelter programs appear to be only modified versions of the self-help housing strategy (Hardiman and Midgley, 1989; Stren, 1990).
Regardless of the frustrating experiences of national urban housing policies in both the capitalist and the socialist Third World, the ideological debate as to how best to house the working poor still remains far from being resolved. Especially, in the socialist Third World, there is an ongoing debate regarding whether or not housing is a social service or just another item of consumption (Mathey, 1990). Following the recent waves of democratisation that swept across Eastern Europe, this debate has even ascended to a higher plane in the socialist African states. Indeed, one of the major challenges that presently face the policy makers of the socialist states of Africa is that of going all the way to privatise the land that is currently under public control.
Although substantial arguments could be made in favour of privatisation, the extent to which such a policy change will improve the housing plight of the working poor is not clearly known. Nowhere in Africa is this question currently demanding as urgent a response as in Ethiopia, which after a change of government in May 1991, still finds itself at the cross-roads between privatisation and continued public ownership of urban land and rental houses.
Ethiopia, like most other Third World nations, had not shown any serious concern over her fast-worsening urban housing problems until very recently. In fact, the country had no urban housing policy to speak of until after the outbreak of the February 1974 Revolution, which replaced the semi-feudal regime of Emperor Haile Selassie with that of the Marxist military junta, i.e., the Provisional Military Administrative Council.
The interest of the Ethiopian government in meaningfully influencing urban housing actually predates February 1974. For instance, the Department of Housing was established within the Ministry of Public Works in 1959. Not surprisingly, therefore, the five-year national economic development plans that were issued following the establishment of this department have successively attempted to indicate the needs for, and to roughly chart the preferred ways of intervening in the urban housing sector. A number of attempts have also been made by the pre-revolution government to prepare master plans for Addis Ababa as well as for other major cities of the country. However, until the outbreak of the February 1974 Revolution, decisions regarding housing permits and building codes were made in most urban centres of the country largely based on local needs and abilities and on vaguely defined and inadequately documented zoning ordinances. It was very difficult for the policy makers of those days to realise the potential damage of such a practice. This was because at that time there was no serious housing shortage in the country, largely owing to the fact that the fast- growing demand for housing continued to be met through increased production of poorly built but seemingly affordable quasi-legal rental accommodation. For instance, out of the 4600 houses that were constructed in Addis Ababa during the 12 months preceding the housing census of 1967, units that had municipal permits accounted for less than 11 percent (Central Statistical Office Dec.1968). Besides, the landlords continued to subdivide or make extensions to existing rental accommodations, often without any supervision by municipal authorities.
Although the gap between the number of housing units and the number of households of cities like Addis Ababa was not an alarming one in 1974, decades of uncontrolled production of substandard and predominantly small rental dwellings had already produced a significant class of tenants who were extremely dissatisfied not only with their poorly-serviced dwellings and neighbourhoods but also with the ways in which a good proportion of their landlords treated them. Thus, following the nationalisation of all rural land in February 1975, thousands of people marched through the main streets of cities like Addis Ababa calling upon the revolutionary government to proceed with the nationalisation of urban land as well. The inexperienced Marxist government, in a desperate attempt to win popularity, promulgated Proclamation No. 47 in July 1975 and thereby nationalised all urban land and rental dwellings in the nation. The proclamation specified that land was no longer to be sold or bought but to be freely provided to needy households. Accordingly, it included provisions that entitled every legally qualifying person or household to the exclusive use of up to 500 square meters of urban land for residential purposes. It also stipulated that a given household could own only a single unit of dwelling.
The Proclamation did not only make rental dwellings public property but also introduced a significant reduction in the monthly rents. This reduction was primarily intended to alleviate the rent burdens of low-income tenants. Accordingly, it spelled out varying levels of rent cutbacks ranging from 50 percent for houses with the lowest monthly rents to 15 percent for houses that were rented for Birr1 300 per month.
Proclamation No. 47/1975 was later followed by a series of additional proclamations and legal notices concerning urban land and housing development. Most of the additional proclamations and legal notices were intended to address the new issues that arose as a result of the public ownership of urban land and rental houses. As such some of them had provisions that superseded some of the stipulations of Proclamation No. 47. Nonetheless, one common thread ran through most of the changing policy measures that were successively adopted during the late 70s and 80s. This common thread was the military junta's declared commitment to laying down all the necessary rules and regulations that would help build the type of socialist economy that it hoped to create in the country.
In general having made all urban land public property and after establishing the necessary institutions that would implement its revolutionary housing policies, the Marxist government felt that it was very well equipped to promote low-income home ownership as never before. It hoped to do this not only by supplying free building lots but also by providing house plans and technical advice free of charge to all qualifying applicants. Like most other governments elsewhere in Africa, it launched various types of self-help housing programs anticipating that such an undertaking would speed up the production of low-income housing.
The research that has so far been done on the performance of Ethiopia's post- revolution housing policies and programs is woefully inadequate. Among these, Koehn and Koehn (1979) predicted that the country's national urban development programs were from the very beginning bound to fail since the Marxist government operated "on the principles of planning for rather than by the urban masses". The results of some recent studies indicate that this prediction has been correct. (Ministry of Urban Development and Housing; 1989; Girma Kebbede, 1992; Gossen, et.al. 1992). Nonetheless, a few writers have asserted that regardless of many shortcomings in both policy and practice, the nationalisation of urban land in Ethiopia has served to promote low-income home ownership (Wendt, et.al., 1990, Asiama; 1990). To what extent is this assertion acceptable? An attempt is made below to address this and other related questions with the help of data that were gathered through a structured questionnaire survey of the employees of the St. George Brewery in Addis Ababa in the summer of 1993.
1 At the time of writing, the official exchange rate was Birr 7.15 for one US dollar.
The St. George Brewery, which was established nearly 80 years ago, is located in one of the busiest hubs of South-western Addis Ababa and had a total of 715 employees at the time of the survey. The employees of this brewery were chosen for the study mainly based on the understanding that such an old factory comprises a substantial pool of low and moderate income workers who have stayed in service long enough to make meaningful tenure choices.
The survey covered a total of 200 employees between the ages of 21 to 55 years and engaged in production and production-related work in the factory. Their mean and median ages were both 36 years and males accounted for 86 percent of the total. About 75.5 percent of the employees were married. Their mean and median monthly family incomes were Birr 365 and 264 respectively. Such a family income structure places most of these workers among approximately the lowest 60 percent of the total households in the city. Nonetheless, it is worth noting that the survey was conducted shortly after the Transitional Government of Ethiopia substantially raised the monthly salaries of the lowest-paid government employees, including those working at St. George Brewery. As regards work experiences the respondents had stayed in service from 2 to 35 years, their mean years of service being 15 years.
With respect to housing tenure, about 20.5 percent of the 200 respondents were homeowners. The corresponding figure for the whole of Addis Ababa during the given period was about 33.0 percent. Among the respondents, public sector renters accounted for 49 percent while tenants in the private rental sector accounted for 22 percent. The remaining 9 percent lived with their parents or relatives. This means that roughly one-third of the respondents had been unable to gain access to public housing.
As regards aspiration to own homes, the overwhelming majority of the non home-owning respondents, about 81.8 percent reported that they were considering either to build or to buy homes at the time of the survey. Even though reported aspirations were high, real attempts made to secure building lots were not as impressive. All in all, only 50.5 percent of the 200 respondents ventured to go beyond aspiration to look for building lots. The proportion of persons that applied to be granted building lots was as low as 45 percent of those workers who did not own homes at the time of the survey.
To examine the extent to which interest in home ownership was related to the awareness of the advantages of public ownership of land, respondents reporting aspirations to own homes were asked whether or not they would refrain from building homes if urban land were privatised. About 55 percent said that they would abandon their plans about building or buying homes if this happened. When workers who were owner-occupiers were asked about the impact of public ownership of urban land on their tenure status, 92 percent reported that they would have remained tenants had it not been for the supply of building lots free of charge.
The main question that needs to be raised at this point is: if the levels of home ownership aspirations among low and moderate income workers were as high as reported by these factory workers, why did one half of them file applications to secure building lots while the other half did not? It appears that two main factors, one internal and the other external, were responsible for keeping down the number of applicants regardless of the availability of land free of charge. The internal factors are those related to the demand-side of the equation and have much to do with household demographic, social and economic characteristics. The external factors tend to be supply-side constraints and as such have much to do with the practical availability of land and credit as well as with the relevant public policy ingredients that either facilitate or impair housing production.
The results of the questionnaire survey yielded adequate data that could be systematically analysed to assess the extent to which the demographic, social and economic backgrounds of the respondents might have made a difference as regards the decision to apply for building lots. To conduct such an analysis, first the relevant data on all respondents who were non home-owners at the time of the survey were extracted. Then with the help of Discriminant analysis, an attempt was made to identify those internal factors that were strongly associated with real aspirations to own homes as expressed through the decision to apply for vacant land.
The main reason why Discriminant analysis was employed in this study is the fact that it is "essentially an adaptation of the regression analysis technique designed specifically for situations in which the criterion variable is qualitative rather than quantitative in nature". It is not uncommon for qualitative variables to be dichotomous in nature as, for instance, in the cases of voters versus non-voters, smokers versus non-smokers, or users versus non-users of a certain agricultural input. In most cases, researchers are interested in quantitative predictor variables that are related to the two criterion values. There are also situations in which a qualitative criterion variable may have more than two values. For example, a researcher may be interested in the differences between Christians, Moslems and Jews or between members of five different ethnic groups. In each case, we may use Discriminant analysis to know "whether values on various predictor variables are related to the alternative values on the qualitative criterion variable" (Kachigan, 1991). In this study, our criterion variable, filing applications to secure land, is dichotomous in nature because our respondents fall into any one of the two categories of applicants and non-applicants.
As provided by Johnston (1984), the basic equation employed in Discriminant analysis is of the following general form:
Xo = f(DI, DII - Dm),
where DI - Dm are the Discriminant functions;
Xo is the dependent variable (the categories into which the observations are divided), m is the number of categories into which the dependent variable is dived less, one.
The variables used in the analysis were the following.
The dependent variable is APLYLN Based on their reactions to the question of whether or not they have made formal applications in order to secure building lots, the respondents were divided into two main categories:
Group 1: if " Yes", and Group 2: if "No".
The independent variables used in the analysis are the following:
1. AGE - respondent's age
2. SEX - respondent's sex: 1, if male; 2, if female
3. MARITAL - marital status: 1, if married; 0, if single
4. EDUC - highest level of education attained
5. PBIRTH - place of birth: 1, if urban; 2, if rural
6. RGBIRTH - administrative region in which the respondent was born: 1, if Addis Ababa; 2, if Shewa (the central region in which Addis Ababa is located); 3, if any other region
7. LCRESID - length of continuous residence in Addis Ababa (years)
8. SATIS - level of satisfaction in the qualities of current dwelling: 1, if high; 2, if moderate; 3, if low
9. LTDWEL - length of continuous dwelling in current residence (years)
10. TENURE - respondent's tenure: 1, if public sector renter; 2, if private sector renter, lodger, caretaker and hostel or hotel dweller
11. OPBBUY - attitude towards home ownership: 1, if considering home ownership as essential for residential satisfaction; 2, if not
12. PTOWNAA - tenure status of respondent's parents if living (or have ever lived) in Addis Ababa: 1, if owning a home; 2, if not
13. HHINC - total household income
14. AAINCGR - average annual income growth
15. NWADULT - number of working adults in the household
16. HHSIZE - household size
17. NCHILD - number of children
18. ASIST - respondent's status regarding regular financial assistance for dependants other than his/her own children: 1, if assisting; 2, if not
19. YRSRVD - total length of period the respondent spent as a gainful worker
20. APLYRNT - the respondent's reaction to the question whether or not he/she had applied to secure a public sector rental accommodation: 1, if "yes"; 2, if "no"
The variables listed above have been selected for inclusion in the analysis because each one of them is expected to influence considerably the respondent's attitude towards home ownership. For instance, the relationship between AGE and home ownership is such that it is the middle-aged individuals rather than either the young adults or the elderly who tend to have both a stronger aspiration for and the economic means to become home owners and to stay in that tenure category. The variables such as SEX, marital status (MARITAL) and education (EDUC) were included based on the assumption that it is generally males rather than females, the married rather than the single, and the more educated rather than those with minimum education who tend to have greater aspiration for home ownership. As regards place of birth (PBIRTH), our expectation is that it is people who were born in rural areas that tend to have stronger interest in home ownership than those born in cities. Region of birth (RGBIRTH) was added to the analysis to examine the extent to which people who were born in areas closer to Addis Ababa have greater chances to buy or build homes in the city in comparison with those who were born in more distant regions. Length of continuous residence in Addis Ababa (LCRESID) is expected to have a positive influence on the probability of home ownership. Comparatively, higher levels of satisfaction (SATIS) in rental dwellings and longer duration of time spent as a tenant (LTDWEL) tend to lower a person's interest in filing applications to secure vacant land. Nonetheless, further examination of tenure as an independent variable suggests that people living in public housing may have a lesser interest in home ownership than those living in private rental homes mainly because the former pay considerably lower monthly rents than the latter for dwelling units of more or less the same size and quality.
Attitude towards home ownership (OPBBUY) was included in the analysis mainly to assess the extent to which an individual's perception of the advantages of home ownership can influence his or her decision with respect to filing applications to secure building lots. In general, it is expected that it is those people who have high a regard for home ownership who would be most inclined to file applications. Tenure status of the respondent's parents in Addis Ababa (PTOWNAA) was added because it was felt that parental tenure might considerably influence a respondent's attitude towards home ownership. For instance, those respondents who expect to inherit property from their parents may be more reluctant to seek vacant building lots than those who have no such expectations.
The variables such as total household income (HHINC), average annual income growth (AAINCGR) and number of working adults (NWADULT) were included because it was expected that an increase in each one of them would tend to raise one's interest in filing applications to secure vacant land. Comparatively, it was expected that a large household size (HHSIZE) would tend to lower the probability of filing applications to secure a building lot. Number of children (NCHILD) was entered in the analysis as an independent variable because a large household size is not necessarily a result of a large number of children. The respondent's status with respect to providing regular financial assistance for dependants other than his own children (ASSIST) was added to the list of independent variables because neither household size nor number of children alone reflects the financial constraint that a home ownership aspirant may face as a result of the assistance that is required of him or her by relatives living elsewhere. The assumption here, therefore, is that it is those people who have no or fewer dependants that are more likely to become home owners than those who regularly provide a significant amount of financial assistance to their relatives. In the analysis, the importance of length of time that a respondent has spent as a gainful worker (YRSRVD) lies in the fact that it is those respondents who had been in service for a longer number of years than those who had fewer number of years as gainful workers who tend to file applications to secure vacant land. The respondent's experience regarding filing applications to secure public sector rental accommodation (APLYRNT) was included in the analysis based on the assumption that those people who had applied for public housing were least likely to refrain from seeking vacant building lots than those who did not.
Initially the cases of a total of 159 non-home-owning respondents were considered for the analysis. Since 15 of these had at least one missing discriminating variable, only 144 cases were used in the analysis. Of these, 65 were applicants while 79 were non-applicants. After running the analysis, some very interesting results were obtained.
First, a brief glance at the group means is in order. As indicated in Appendix 1, the group means suggest that in a striking similarity to the case of most developed countries, tenure choice in socialist Ethiopia was largely dependent on life cycle. In general, workers who filed applications to obtain building land tended to be older, married, more educated and wealthier than those who did not. The findings also imply that applicants generally had fewer children and lesser number of dependants than non-applicants. The group means also suggest that those who applied were mostly rural born in-migrants that had spent fewer years of continuous residence in the city in comparison with those who did not apply. Further examination of the group means indicates even more interesting results. It suggests that applicants came largely from the apparently large pool of public sector renters who were the most dissatisfied with their dwellings.
Even though the group means have yielded interesting results, they do not of themselves indicate the variables that were most important in discriminating between applicants and non-applicants. In fact, most of the differences in the group means are not necessarily statistically significant and as such their relevance does not extend beyond the mere suggestion of the most likely scenarios. Thus, in order to identify those variables that discriminate the most between the two groups, we have to examine the results of the standardised canonical Discriminant function.
Table 1. Canonical Discriminant Functions
Pct of Cum Canonical After Wilks'
Fcn Eigenvalue Variance Pct Corr Fcn Lambda Chisquare DF Sig
: 0 .6844 51.198 14 .0000
1* .4612 100 100 .5618 :
* marks the 1 canonical Discriminant function remaining in the analysis.
Table 2. Classification Results
No. of Predicted Group Membership
Actual Group Cases 1 2
Group 1 65 51 14
78.5% 21.5%
Group 2 79 21 58
26.6% 73.4%
Percent of "grouped" cases correctly classifie 75.69%
As shown in Table 1, the canonical Discriminant function in this analysis has a correlation coefficient of .5618, which is significant at the 0.0000 level. As indicated in Table 2, the level of classification achieved with this function was such that the group membership of 78.5 percent of the cases in Group 1 and 73.4 percent of those in Group 2 have been correctly predicted. All in all, 75.69 percent of the grouped cases were correctly classified. This means that the results of the canonical Discriminant function are good enough to be used for the identification of those variables that discriminate the most between the two groups. The coefficients of the standardised canonical function are shown in Table 3. As regards the structure matrix, however, it is not shown here since it does not carry high values in any one of the variables and is as such difficult to name.
Table 3. Standardised Canonical Discriminant Function Coefficients
FUNC 1
MARITAL .22592
RGBIRTH .35098
LCRESID .31718
LTDWEL -.53424
SATIS .60717
NCHILD -.76415
OPBBUY .34064
HHINC .56348
HHSIZE .41604
ASIST -.17274
PTOWNAA -.25776
AAINCGR .17406
TENURE -.56746
APLYRNT -.35093
Table 3 indicates that those four variables, namely NCHILD, SATIS, TENURE and HHINC in order of importance, have the highest standardised canonical Discriminant function coefficients. This means that it was mostly public sector renters with fewer children, lowest residential satisfaction and larger household incomes that were motivated to file applications to obtain building lots.
Three important conclusions can be drawn from the above findings. First, the high values of NCHILD and HHINC indicate the significance of the stages in the workers' life cycle as regards the decision to file applications to secure building land. It means that it was those workers who had fewer children but higher incomes who had the greatest tendency to apply. Second, the high coefficient of SATIS suggests that it was not necessarily renting per se that people were running away from when they attempted to become homeowners. In post-revolution Addis Ababa, it has apparently been easier to secure vacant building lots than unoccupied rental homes. Among the workers interviewed at the St. George Brewery, the proportion of total applicants (both homeowners and renters) who were able to secure land by August 1993 were 41.8 percent. In contrast, among the persons who had been seeking affordable rental accommodation, the proportion that was able to find public sector rental homes at the time of the survey was only 20.5 percent.
In view of the facts mentioned above, and also given the daunting challenges of building one's own home in a city like Addis Ababa, it appears that the level of residential satisfaction that the workers drew from their accommodation did play a significant role as regards the decision to seek vacant building lots. Put differently, this means that some of the respondents who had applied to secure building land might have reconsidered their plans had they been provided with affordable rental accommodation During the survey, those respondents who reported interest in home ownership were asked two questions that were directly relevant to this issue. First, they were asked if they had any plan to rent better accommodation. About 50 percent answered in the affirmative. Secondly, they were asked if they would give up their plans should they find affordable rental accommodation. About 35.2 percent reported that they would abandon their plans to build or buy homes should they find affordable rental homes.
Thirdly, the high coefficient of TENURE indicates that irrespective of the scarcity of public sector rental homes, a substantial proportion of the respondents preferred to exhaust the search for such an accommodation before grappling with the greater idea of home ownership. Especially it appears that most of the workers who at the time of the survey lived with their parents or relatives and those with various types of informal living arrangements lacked the psychological as well as the financial readiness to solve their housing problems by building new homes.
There were at least two major external factors that discouraged low and moderate income workers from filing applications to secure building land. These factors were the low and misdirected investment in housing development and the bureaucratic red tape in the land and credit procurement process.
Regarding the first factor, investment in housing as percentage of gross domestic product in Ethiopia has remained well under 3 percent for the greater part of the last two decades (Gossen, et. al. 1992). The principal factor behind such a low rate of investment appears to be, on the one hand, the economic downturn that the country has witnessed since the mid 70s and, regions on the other hand. the drain caused on its human and financial resources by the protracted civil war in its northern region.
To make matters worse, the limited funds that were allocated for housing development were used in ways that, either by design or default, overlooked the housing needs of the low and moderate-income households. In the first place, approximately 50 percent of all the loans approved for housing development by the Housing and Savings Bank went to the state housing sector mainly to finance the construction of middle and upper-income dwellings and various types of office complexes. Secondly, most of the funds that were allocated for housing co-operatives and individual applicants were used to meet the housing needs of people in the upper income bracket. For instance, even though households with monthly earnings of Birr 500 or more constituted only about 15 percent of the total households in the city (Ministry of Urban Development and Housing, 1989), they received over 65 percent of all the funds that were made available for housing co-operatives and individual applicants between 1975 and 1990. In contrast, although households with monthly incomes of Birr 300 or less accounted for nearly three-fourths of the total households in the city, their share of the loans that were approved by the bank during the given period was only 10.4 percent (Housing and Savings Bank, 1990).
One major implication of the foregoing discussion is that it is mostly those low and moderate income aspirants who could secure funding from sources other than the public sector that were encouraged to apply for land. In fact, the survey results indicated that most of the respondents had a clear awareness of their inability to depend on the then existing system of funding housing. For instance, 79.2 percent of the home-owning respondents depended on funds obtained from non-governmental sources while 54 percent of those who expressed interest in home ownership reported that they were anticipating to use either their own savings or funds they might obtain from relatives and friends.
As regards the bureaucratic red tape in the land and credit acquisition process, there seems to be little or no doubt that it was one of the major obstacles that people aspiring to become home owners in Addis Ababa had to face irrespective of their income background. This happened largely due to the fact that there were so many institutions involved in the land allocation process that it took from several months to several years before all had said and done all that was required to assign a building lot to a prospective home owner.
In general, it seems that the bureaucratic hurdles in the land acquisition process had two important effects on the development of low-income housing in Addis Ababa. In the first place, it appears that due to the unduly lengthy land acquisition processing period, many of the applicants simply become disheartened and abandoned their plans for building homes. Moreover, even if their patience withstood the red tape, the long waiting period normally exposed a substantial number of them to another detrimental factor: loss of status as regards qualification for public sector credit either due to ageing or as a result of the fast-rising cost of building materials.
To add to the disappointment of low-income applicants, even if they finally succeeded in acquiring building land, the assigned lots were most likely to be located in peripheral and undeveloped quarters of the city (Gossen, et. al., 1992). The survey results in fact showed that over 80 percent of those respondents who managed to become owner-occupiers were living in peripheral neighbourhoods. This is mainly because the zoning regulations did not encourage the construction of low-cost houses in more internal locations. There is hardly any doubt that such residential locations exposed the prospective homeowners to burdensome land development and construction costs as well as to greater commuting distances.
The second major impact of the shortcomings in the land allocation process was that the example of both the partly successful and the totally disappointed applicants could considerably deter prospective applicants from going ahead with their plans. The results of the questionnaire survey have yielded limited but valuable insights in this respect. Among the workers who were not homeowners, only 11 of the 72 who filed applications were able to secure building lots at the time of survey. The period those workers spent waiting for lot assignment ranged from 2 to 8 years, their mean waiting time being 5.73 years. Of the 11 recipients of land, only one person was able to start housing construction at the time of the survey.
Most of the non-home-owning workers of the St. George Brewery in Addis Ababa were interested in becoming owner-occupiers. However, at the time of the survey, half of the persons reporting interest in home ownership had not filed any applications to secure building lots. Apparently, the decision to file an application was much more a function of the stages in the workers' life cycle and their levels of satisfaction in the quality of their dwellings than of their awareness of the availability of free land. This should not be surprising since in post-revolution Addis Ababa, it had apparently been easier to secure vacant building lots than to obtain unoccupied rental homes. Indeed, the results of the study indicate that more than one-third of the workers who were considering to build their own homes at the time of the survey would easily have settle for affordable rental accommodations if they had been given a choice.
In general, the experience of the workers of the St. George Brewery in Addis Ababa shows that public ownership of land has considerably improved the home ownership affordability status of many low and moderate-income workers. The number of homeowners it has created are, however, far fewer than the number of home ownership aspirants that it has helped to generate. Regardless of this fact, the findings of this study have not conclusively shown that urban land needs to be privatised. Nonetheless, they have furnished enough hints suggesting that in present day Addis Ababa, the moral and political rationale of public ownership of land clearly outweigh its economic justifications. Of course, one cannot deny the fact that the prospect of increasing low-income home ownership rests to some extent in some form of public control of urban land. However, the fact remains that given the current social and economic state of affairs in Addis Ababa, it will be almost exclusively the middle and upper-income households who will continue to benefit from public ownership of land in the foreseeable future.
Asiama, Seth Opuni. 1990. "Land For Housing the Urban Poor in Africa - Some Policy Options" in Philip Amis and Peter Loyd (eds.), Housing Africa's Urban Poor: International African Seminars. Manchester: Manchester Univ. Press, 239-251.
Burgess, Rod. 1979. "Petty Commodity Housing or Dweller Control? A Critique of John Turner's Views on Housing Policy" in Ray Bromley (ed.), The Urban Informal Sector: Critical Perspectives on Employment and Housing Policies. Oxfor Pergamon, 1105-1133.
Central Statistical Office. Dec. 1968. Survey of Major Towns in Ethiopia.
Devas, Nick. 1983. "Financing Urban Land Development for Low-income Housing: An Analysis with Particular Reference to Jakarta, Indonesia" in Third World Planning Review. 5, 3: 209-225.
Gossen, F., et.al. 1992. "From Sources to Policies: Institutional vs. Geographical Models for Evaluating Housing Conditions in Addis Ababa". An unpublished paper prepared for the Second National Conference of Ethiopian Studies, Addis Ababa, October 5-7, 1992.
Hardiman, M. and J. Midgley. 1989. The Social Dimensions of Development. London: Gower.
Housing and Savings Bank, 1990. Annual Report.
Jenkins, Paul. 1990. "Mozambique" in Kostha Mathey (ed.), Housing Policies in the Socialist Third World. London: Mansell, 147-179.
Johnston, R. J. 1984. Multivariate Statistical Analysis in Geography. New York: Longman.
Kachigan, Sam Kash. 1991. Multivariate Statistical Analysis: A Conceptual Introduction. New York : Radius Press.
Kebbede, Girma. 1992. The State and Development in Ethiopia. London: Humanities Press.
Khabagambe, Denis and Cliff Moughtin. 1983. "Housing the poor: a case study in Nairobi" in Third World Planning Review. 5, 3: 227-248.
Koehn, Peter and Eftychia Koehn. 1979. "Urbanisation and Urban Development Planning in Ethiopia" in R.A Obudho and Salah El-Shakhs (eds.), Development of Urban Systems in Africa. New York: Praeger.
Lea, John. 1980. "Self-help and autonomy in housing: theoretical critics and empirical investigators" in Murrison, Hamish S. and John P. Lea (eds.), Housing in Third World Countries: Perspectives on Policy and Practice. New York: St. Martin's Press, 49-53.
Mabogunje, Akin L. 1990. "Urban Planning and the Post-colonial State in Africa: A Research Overview" in African Studies Review. 33, 2: 121-203.
Macoloo, Gervas Chris. 1988. "Housing the Urban Poor: A Case Study of Kisumu Town, Kenya" in Third World Planning Review. 10, 2:159-174
Main, H. A C. 1990. "Housing Problems and Squatting Solutions in Metropolitan Kano" in Robert Potter B. and Ademola T. Salau, (eds.), Cities and Development in the Third World. London: Mansell, 12-31.
Mathey, Kostha. 1990. "Socialist Housing: Some Key Issues" in Kostha Mathey, (ed.,) Housing Policies in the Socialist Third World. London: Mansell, 14-31.
Ministry of Urban Development and Housing. 1989. Housing: Addis Ababa. Report Number 3: Final Report.
Nientied, Peter and Jan van der Linden. 1985." Approaches to Low-income Housing in the Third Worl Some Comments" in International Journal of Urban and Regional Research. 9, 3:311-329.
Pezzoli, Keith. 1987. "The Urban Land Problem and Popular Sector Housing Development in Mexico City" in Environment and Behaviour. 19, 3: 371-397.
Salau, Ademola T. 1990. "Urbanization, Housing and Social Services in Nigeria: The Challenge of Meeting Basic Needs" in Robert B. Potter and Ademola T. Salau, (eds.), Cities and Development in the Third World. London: Mansell, 51-69.
Stren, Richard E. 1990. "Urban Housing in Africa: The Changing Role of Government Policy" in P. Amis and P. Loyd (eds.), Housing Africa's Urban Poor, Manchester: Manchester University Press, 35-54.
Turner, John F. C. 1976. Housing by People: Toward Autonomy in Building Environments. London: Marion Boyars
_______. 1979. "Housing in Three Dimensions: Terms of Reference for the Housing Question Redefined" in Ray Bromley (ed.), The Urban Informal Sector: Critical Perspectives on Employment and Housing Policies. Oxfor Pergamon, 1135-1145.
Wendt, Sabine et.al. 1990. "Ethiopia" in Kostha Mathey (ed.), Housing Policies in the Socialist Third World. London: Mansell, 227-248.
Group means
APLYLND AGE SEX MARITAL PBIRTH
1 34.92308 1.15385 .73846 1.80000
2 34.51899 1.12658 .68354 1.74684
Total 34.70139 1.13889 .70833 1.77083
APLYLND RGBIRTH SATIS EDUC YRSRVD
1 2.40000 2.72308 3.90769 14.52692
2 2.12658 2.43038 3.68354 13.04051
Total 2.25000 2.56250 3.78472 13.71146
APLYLND HHSIZE OPBBUY LCRESID LTDWEL
1 5.36923 1.38462 21.41154 11.36738
2 6.02532 1.20253 21.59494 13.74557
Total 5.72917 1.28472 21.51215 12.67208
APLYLND PTOWNAA HHINC AAINCGR NWADULT
1 1.90769 364.69231 32.25749 1.30769
2 1.87342 260.94937 29.91216 1.31646
Total 1.88889 307.77778 30.97082 1.31250
APLYLND ASIST NCHILD TENURE APLYRNT
1. 64615 2.47692 2.32308 1.40000
2 1.79747 3.12658 2.36709 1.63291
Total 1.72917 2.83333 2.34722 1.52778
Group Standard Deviations
APLYLND AGE SEX MARITAL PBIRTH
1 6.67155 .36361 .44289 .40311
2 7.01438 .33463 .46806 .43760
Total 6.84093 .34704 .45612 .42176
APLYLND RGBIRTH SATIS EDUC YRSRVD
1 .70267 .54508 1.15546 5.9253
2 .70469 .72825 1.18261 8.51583
Total .71449 .66605 1.17171 7.47130
APLYLND HHSIZE OPBBUY LCRESID LTDWEL
1 2.54073 .49029 8.57167 7.96040
2 2.81012 .40445 9.54143 8.85671
Total 2.70255 .45286 9.08566 8.51804
APLYLND PTOWNAA HHINC AAINCGR NWADULT
1 .29171 273.80011 43.11821 .49759
2 .33463 117.60864 21.83354 .67071
Total .31537 209.23649 33.06770 .59683
APLYLND ASIST NCHILD TENURE APLYRNT
1 .48188 2.27148 .53349 .49371
2 .40445 2.82556 .48509 .48509
Total .44594 2.60177 .50618 .50097