Abstract

It is now widely recognised that low-income households in Third World cities often use their homes not only as a means of shelter, but also as a source of income. This paper focuses on the types of enterprises found in five low-income neighbourhoods in the intermediate sized city of Pereira, Colombia. It is argued that it is necessary not only to recognise the importance of home-based enterprises for the economies of low-income households, but also to consider how these enterprises vary between households and over time. As the nature of both individual household circumstances and of neighbourhoods alter over time, so do the types and number of home-based enterprises operating.

Keywords

Home-based enterprises, self-help housing, renting, Colombia

Katherine V. Gough: Institute of Geography, University of Copenhagen,
Øster Voldgade 10, 1350 Copenhagen K., Denmark.

Danish Journal of Geography 96: 95-102, 1996.

For many low-income households the sole function of their home is to provide shelter, but some houses also become a source of income as home-based enterprises are established or rooms are rented out. In almost all low-income settlements in Third World countries people can be seen cooking, sewing, selling drinks and food, keeping animals, mending and making shoes, manufacturing various goods, cutting hair, giving injections, keeping animals and renting rooms. Several authors have lamented that the use of the house as a source of income is a topic which has been neglected despite the important role that home enterprises play in the economic strategies of many low-income families (McCallum and Benjamin, 1985; Strassmann, 1987; Gilbert, 1988). There have, however, been some anthropological studies on the use of the home as a workplace in low-income settlements (Peattie, 1968; Perlmann, 1976; Lobo, 1982; Bohman, 1984). These authors have provided clear evidence of the prevalence of the use of the home as a workplace, especially by women, and the importance of these activities in the budgets of individual households. The studies have tended, though, to focus on individual households, rather than on the frequency of home-based enterprises across the city (Gilbert, 1988).

UNCHS/ILO (1995) have recently provided a summary

of the existing literature on the use of shelter as a work place. They claim that the chief advantage of home-based enterprises is the contribution which they make to the household economy of low-income people. As the premises used for the 'business' also provide a place to live, the overheads are very low resulting in lower marginal capital investments for setting up an operation in the home than elsewhere. The disadvantages associated with homebased enterprises are claimed to be: first, the isolation and lack of visibility in which the workers carry out their tasks which permits greater levels of exploitation than when the work is carried out in factories; and second, the effect of economic activities on the residential environment in both monetary and environmental terms. Attempts to weigh up the relative costs and benefits of home enterprises, however, tend to come out in favour of allowing them (Gilbert, 1988; Tipple, 1993).

This paper provides a city-wide account of the homebased enterprises operating in low-income barrios in the intermediate sized city of Pereira, Colombia. The range and frequency of home-based enterprises operating and the use of the home as a means of earning rental income are discussed. The ways in which these enterprises vary according to the nature of the household, the house and the settlement type are analysed. (1)

The Study Area

Since the 19305, Colombia has undergone a transition from a rural society to a predominantly urban one. The proportion of the population living in urban centres increased from 31% in 1938 to 65.3% in 1985. Unlike other Latin American countries, this transition has not resulted in the emergence of one primate city, but in a network of large, intermediate and small urban centres. However, as in many Latin American countries, the housing supply has failed to keep up with demand and the urban poor have been unable to afford the housing built by the construction industry. This has resulted in a proliferation of self-help housing in illegal settlements and in state sites-and-services projects.(2)

Since its establishment in 1863, Pereira has grown from a small, subsistence based hamlet to an urban conurbation housing almost 350,000 people and a centre for the coffee growing and textile industry. The city has developed clearly demarcated sectors with the central area the focus for commerce, services, industry, middle and high-income housing, and rooms rented by the poor. Most of the lowincome owner-occupier households live out on the periphery of Pereira in an area referred to as Cuba, or in Dosquebradas (figure 1).

This research was carried out in five low-income settlements in Pereira of varying legality and age. 2,500 lotes was a new state sites-and-services project where households were in the early stages of construction, and Villa Santana was the most recent pirate (illegal) settlement in Pereira where the houses were not more than 5 years old. Three older settlements aged between 8 and 10 years were also studied: El Plumön originated as an invasion along a disused railway line, Leningrado was founded by a selfhelp housing association, and Las Mercedes was a state sites-and-services project.

A formal questionnaire survey was conducted personally
by the author with 250 randomly selected households in

the five self-help settlements. In the three older barrios a 20% sample was employed resulting in 30 questionnaires being conducted in El Plumön, 35 in Las Mercedes, and 40 in Leningrado. In Villa Santana a 10% sample was used and 95 interviews were conducted. In 2.500 lotes, a 15% sample of resident households, 50 in total, were questioned.(3) The main focus of the research project was to investigate the construction and consolidation of self-help housing. As in many cases the houses acted as a source of income as well as providing shelter, data were collected on the various uses of the houses. This paper is based on the data obtained from the questionnaire survey supported by in-depth case studies and participant observation conducted during a period of 12 months.

The two state sites-and-services projects, Las Mercedes and 2,500 lotes, were legal and had service installations from the start. All the houses in Las Mercedes had been connected to the mains sewerage, water and electricity supply but in 2,500 lotes not all the households could afford to make the necessary connection to the mains: 16% were using a neighbours toilet, one-third had an outdoor tap in their plot, and a few had made an illegal connection to the electricity supply.(4) El Plumön, Leningrado and Villa Santana all originated as illegal settlements. In 1986, Leningrado had become legal, a legalisation programme for Villa Santana was underway and was being discussed for El Plumön. All the houses in El Plumön and Leningrado had all three services but in Villa Santana parts of the settlement had yet to have the water and sewerage mains installed and not all households were able to afford

to connect their houses to the mains. As a result, 18% of households had only an outdoor tap in their plot, and a few families went to a neighbour for water or to a communal tap. Some of those in the higher section of Villa Santana had no toilet and another 17% were not connected to the sewerage system.

The standard of the housing in the settlements is illustrated in table 1, where a score was calculated for each house according to the type and condition of the building materials used. The standard of the building materials found in the houses in 2,500 lotes and Villa Santana are very similar. A little over half of the houses in each settlement were still in temporary materials, some of which were of very poor quality. In the older barrios, few houses were still built out of poor quality materials and between 38% and 52% of houses were in the most consolidated category. The majority of households, therefore, managed to improve the quality of the materials used in their houses over time. As well as gradually improving the quality of the building materials of their dwellings, many households also expanded the area of their house. The ability to consolidate was not übiquitous, however, and some households were still occupying a very small dwelling even after ten years (Gough, 1992).

Housing as Workplace

Type of Home-based Enterprise

A range of home-based enterprises was identified in the

survey settlements, broadly divisible into two categories: manufacturing and retail/commercial (table 2). The most common type of home-based enterprise was retailing which accounted for almost 60% of the activities reported. Retailers were also found to be the most frequent type of enterprise in Bogota (Peattie, 1980; Gilbert and Ward, 1985) and elsewhere (McCallum and Benjamin, 1985; ILO/WB, 1995). General stores selling food, soap products, kerosene etc., were mainly run jointly by couples for whom it was their only source of income. These stores tended to be in the front room of a brick house and hence were only found in the more consolidated houses (figure 2). Their customers were all local people, often children sent by their mothers, who tended to purchase in small

quantities. There was often an almost constant stream of customers, many of whom were purchasing single items such as one onion or one egg. The general stores play an important role in low-income settlements offering credit as well as selling in small quantities, and saving local residents travel time. Three of the general stores questioned also doubled as bars with a pool table. These enterprises were in two storey houses; the household lived on the first floor and the ground floor was devoted to the business.

The sale of drinks and sweets etc. on a smaller scale was found to operate from a wide range of house-types, but these stores were predominantly run by women. One exception was an old man in Leningrado who sold sweets and cigarettes from a trolley in front of his house. He had been forced to stop selling in the centre of Pereira after the introduction of a sales license for street sellers which he could not afford. As a result of having to sell from his home, his sales had dropped with severe consequences for him and his wife for whom it was their only income. There were also two cases where individuals produced as well as sold food: one was a male baker and the other was a woman who made arepas (unleavened bread) and empenadas (meat encased in pastry). Personal services were also on offer including injections, washing and ironing, and hairdressing, all carried out by women.

The manufacturing activities included both production on a piecework basis for a local entrepreneur, and activities where production was on a one-off basis for a local customer or for sale on the street. Four shoe-makers, two tailors and one leatherworker all operated under the putting-out system. These occupations were male dominated and in most cases the work was carried out by two or more male members of the household, although in one instance a young woman worked together with her husband as a tailor. As the workshops filled at least one room, they all operated from relatively large, consolidated houses and tended to form the main source of income for the household. Small, poor quality houses did not have enough space for the equipment needed, and there was often also the possibility of rain leaking through the roof or the bamboo walls, damaging equipment and materials.

There were two examples where production was for the open market. In one enterprise, a man made chandeliers which he sold on the streets in the centre of Pereira, and in another, a woman made dolls which she sold from her home. In both instances the money earned was an importantsource of income. The most common form of manufacturingwithin

facturingwithinthe home, however, was dressmaking done to order. This was carried out entirely by women. For some women, and especially those who were single, such work was an important source of income. For others, it was very much supplementary to the man's income, although it could still be vital especially in times of male unemployment.One woman, for example, reported that the money she earned from dressmaking paid for her children's education.

Frequency of Home-based Enterprises

Peattie (1974) found that 19% of households in a lowincome settlement in Bogota ran enterprises from their home. More generally for Bogota a figure of 12% was reported by Gilbert and Ward (1985), 18% in Medellm (Mesa, 1990) and 20% for Santa Marta (Kellet, 1992). Similarly, Strassmann (1985) discovered home-based

enterprises in 17% of houses in the squatter settlements of Lima. There was evidence in all of the barrios studied in Pereira that some houses were not only used as buildings in which to live, but also as places from which a store was run or a workshop operated. On average, 20% of the households questioned in Pereira had some form of income-earning activity within the home (table 3).

Home-based activities were most common in 2,500 lotes where they operated from 28% of the houses. At first sight this is surprising insofar as it is generally believed that newer settlements have fewer activities (Gilbert, 1988) and because restrictions on the establishment of home-based enterprises are often reported in state projects. Many of the enterprises operating in 2,500 lotes, however, were very small; there were many households who sold drinks, ices and food often only on Sundays when the settlement became a hive of activity with people building their homes, many of whom were not yet resident. If these very small, temporary enterprises are removed from the calculations, 18% of households in 2,500 lotes had home-based enterprises, a figure very similar to that in the other barrios. There was, therefore, no clear trend for the number of enterprises operating to increase with the age of the barrio in Pereira. Neither was there any restriction on the operation of home-based enterprises in the state settlements which would have reduced the number of enterprises in these barrios.

Housing and Rental Income

The importance of renting as a housing strategy has been increasingly recognised in recent years (Rakodi, 1995). Research in various Latin American contexts has revealed that many low-income families rent out rooms within their home (Edwards, 1982; Gilbert, 1983; 1987; 1993; Green, 1987; 1988; Gilbert and Varley, 1991). Within the survey settlements, however, only nine owner-occupier households(4%) rented out rooms (table 4). Renting was absent in the newest settlement 2,500 lotes, and occurred in only one house in Villa Santana. In the three older settlements, between 7% and 9% of households rented out rooms. This is still considerably lower than Gilbert and Ward's (1985) finding that 18% of owner families were earning money from rent in Bogota. The low figure found for Pereira is probably due to the relatively young age of the settlements studied. As settlements are consolidated and houses increasein

creaseinsize, more renters and landlords are found (Peattie, 1980; Gilbert and Ward, 1985). It is also possible that over time, as family members leave the home, more space is made available for renting.

Baross and Mesa (1986) found that rental units in Medellin were located in the better housing categories and suggest that the level of family investment must reach a high level in a settlement before renting becomes a common activity. In the survey settlements in Pereira, all of the houses where rooms were rented out were consolidated brick houses, the majority of which had two storeys. In most of these houses, the owners lived upstairs and rented out the bottom half of the house. This created a formal separation of the space used by the owners and their tenants. In two houses where the renters were also relatives, however, the living space was shared and the rent included board as well as lodging. Rent made up 25% of the total household income on average. Despite the small size of the sample in Pereira, the contribution of money from renting to the household budget was similar to that found by Gilbert and Ward (1985) in Bogota (28%) and Edwards (1982) in Bucaramanga (30%). As Gilbert (1983, 1993) and Green (1987) have observed elsewhere in Latin America, petty landlordism was the norm; most landlords were themselves poor, they rented out a few rooms as a sideline to supplement their income. Renting by low-income households is, therefore, more a survival strategy, and a means of consolidating the house, than a profit-making enterprise (Gilbert, 1988).

In the settlements studied, in 7% of cases the entire house was rented out. The frequency of rented houses varied with the age of the barrio, no houses were rented in 2,500 lotes, 7% were rented in Villa Santana, and between 10 and 12% were rented in the three older settlements (table 4). This coincides with Baross and Mesa's (1986) finding that in the younger settlements in Medellin, only 5-10% houses were rented; only in the older settlements did the renting of houses occur to a greater degree (about 25% in settlements over 20 years old). Income earning from renting is, therefore, strongly associated with the more consolidated houses.

Discussion

There are several issues which arise from these data. First,
a clear divide is identifiable between the kinds of home-

based enterprises run by men and by women. Male-run enterprises tend to be more visible and earn more money. In order to understand why this distinction occurs, it is important to understand 'how cultural definitions of women as wives and mothers intervene in economic processes' (Bohman, 1984: 129). In Colombia, a married woman's place is considered to be in the home; if she has to find paid work, the family loses prestige. If it is necessary for a married woman to work, she will try to find some activity which is compatible with her roles as wife and mother. Bohman (1984) found in Medellin that of women in employment, almost half worked within their own homes especially as dressmakers or in small selling operations. Given Colombian men's attitude to women's work and its invisibility, it is likely that many female income-earning activities go unrecorded, especially when men are interviewed. In my own fieldwork, on two occasions when conducting the questionnaire with male heads of household, they claimed that their wives did not work despite the fact that in one case we were sitting in a hairdressing room and in another in a small store, both of which were clearly being run by their wives. If men fail to report these activities, the question remains as to how often they do not mention other activities, such as dressmaking, which are less readily visible.(s)

Another issue is differences in the life-span of the enterprises. Some of the home-based enterprises were destined to have a short life-span; one example was households who set up a small store as their only way of earning a living after becoming unemployed. These stores tended to stock a limited range of goods given the limited capital available, and were viewed as only being temporary sources of income by their owners. In 2,500 lotes, some households responded to an immediate demand for refreshments on Sundays. In time, these enterprises are likely to cease functioning and be replaced by others, such as general stores, which cater for the longer term needs of the community. Not all the home-based enterprises encountered, however, were of a temporary nature. Some of the workshops appeared to be thriving and their owners were talking of expanding the business. Other enterprises were likely to survive as their owners had no alternative way of making money. This applied especially to women who had been left by their husbands; the only way these women could both earn a living and look after their children was to work from the home.

Most of the people working in home-based enterprises

were not able to state how much they earned since they had no fixed wage. It was often not even possible for them to estimate an average income, since they made little attempt to keep track of purchases and sales, and many households spent the money they earned straight away. It was, therefore, impossible to estimate the average contribution of home-based enterprises to household incomes. Peattie (1980) encountered the same problem in Bogota. It was evident, however, that money earned from home-based enterprises helped households finance the consolidation of their home. In turn, as the houses were consolidated, households had greater possibilities for setting up an income generating activity in their home.

One of the most common complaints of people running home-based enterprises was the inadequacy of services. Especially in the newer barrios, electricity supplies were intermittent and the water supply unreliable. Those with refridgerated goods lost stock when the electricity supply was off for longer periods and some workshops were forced to stand idle. Other enterprises, especially some of those involved in food preparation, were particularly adversely affected when the water supply was cut off. There is, therefore, an urgent need for infrastructural improvements to be made in low-income neighbourhoods which would not only benefit all residents but also enable home-based enterprises to operate more efficiently.

Conclusion

In the survey settlements in Pereira, 20% of households used their home as a workplace. Some of the larger, more consolidated homes housed shops or workshops which provided the main source of income for the family. In other instances, the money earned from the home was a minor, but still important, part of the household income. Especially for some single women, working in their home provided a vital source of employment which they could combine with their domestic responsibilities.

Money earned from the home-based enterprises aided the consolidation process of the houses. Renting was confined to the more consolidated houses, though the renting of rooms was not a common way of earning extra income for the households interviewed, mainly because of the relatively young age of the settlements studied. Hence, as households consolidate their dwellings, the opportunities open to them to earn extra income, either through the rent-

ing out of rooms or the establishment of home-based enterprises,

Home-based enterprises in self-help settlements provide a vital source of income for some households. They also provide important services for the other residents of the low-income neighbourhoods. None of the enterprises encountered in Pereira were damaging to the environment nor were they a source of complaint for their neighbours. This suggests that, as others have argued (Gilbert, 1988; UNCHS/ILO, 1995), home-based enterprises should generally not be restricted in low-income settlements. Moreover, they need to be actively supported by infrastructural improvements which would benefit the low-income settlements as a whole.

---

Notes

(1) This research was funded by an award from the Economic and Social Research Council linked to the work of Professor Alan Gilbert of the Department of Geography, University College London. The bulk of the fieldwork was carried out in 1987.

(2) There is no agreed definition of 'self-help', but it is generally used to cover low-income housing where the owners control the building process and have built some or all of the house themselves. In sites-and-services projects, the state sells serviced plots of land on which households build their own house.

(3) Of the 250 questionnaires conducted 7 had to be abandoned, hence the total number of households cited in the data is 243.

(4) Unless otherwise stated, all data quoted in this paper are basedon the author's household surveys.

(5) In order to try to avoid this potential bias in the data, all those interviewed were questioned in detail about any possible income earning activities which took place in the home. As the majority of questionnaires were, in fact, carried out with women, or jointly with women and men, it is hoped that most activities were recorded.

References

Baross, P. and Mesa, N. (1986): From land markets to housing
markets: the transformation of illegal settlements in Medellin.
Habitat International, 10,3: 153-170.

Bohman, K. (1984): Women of the barrio: class and gender in a
Colombian city. Stockholm studies in social anthropology,
13.

Edwards, MA. (1982): Cities of tenants: renting among the urban poor in Latin America. In: Gilbert, A.G., Hardoy, I.E. and Ramirez, R. (eds.) Urbanization in contemporary Latin America. John Wiley & Sons, Chichester.

Gilbert, A.C. (1983): The tenants of self-help housing: choice or
constraint in the housing market of less developed countries.
Development and Change, 14: 449-477.

Gilbert, A.C. (1987): Latin America's urban poor: shanty dwellers
or renters of rooms? Cities, 4, 1: 43-51.

Gilbert, A.C. (1988): Home enterprises in poor urban settlements:
constraints, potentials, and policy options. Regional
Development Dialogue, 9, 4: 21-37.

Gilbert, A.G. (1993): In search of a home. University College
London Press, London.

Gilbert, A.G. and Varley, A. (1991): Housing the poor in urban
Mexico. London, Routledge.

Gilbert, A.G. and Ward, P.M. (1985): Housing, the state and the
poor: policy and practice in three Latin American cities.
University of Cambridge Press, Cambridge.

Cough, K.V. (1992): From bamboo to bricks: self-help housing
and the building materials industry in urban Colombia.
Unpublished Ph.D. thesis, University of London.

Green, G. (1987): Finding a home in a frontier city: the dynamics
of housing tenure in Santa Cruz, Bolivia. Unpublished
Ph.D. thesis, University of London.

Green, G. (1988): The quest for tranquilidad: paths to home
ownership in Santa Cruz, Bolivia. Bulletin of Latin American
Research. 7, 1: 1-15.

Kellet, P. (1992): Residential mobility and consolidation processes
in spontaneous settlements: the case of Santa Marta,
Colombia. Third World Planning Review, 14, 4.

Lobo, S. (1982): A home of my own: social organisation in the
squatter settlements of Lima, Peru. University of Arizona
Press, Tuscon.

McCallwn, D. and Benjamin, S. (1985): Low-income urban
housing in the Third World: broadening the economic perspective.
Urban Studies, 22: 277-287.

Mesa, N.E. (1990): Economic use of housing in 'pirata' and invasion 'barrios', in: Medellin in Raj, M. and Nientied, P. (eds.) Housing and income in Third World urban development, Aspect Publishing, London.

Peattie, L (1968): The view from the barrio. University of
Michigan Press, Ann Arbor.

Peattie, L (1974): The concept of marginality as applied to
squatter settlements. Latin American Urban Research, 4: 101-109.

Peattie, L (1980): Anthropological perspectives on the concepts of dualism, the informal sector, and marginality in developing urban economies. International Regional Science Review, 5, 1: 1-31.

Perlman, J. (1976): The myth of marginality: urban poverty and
in Rio de Janeiro. Univ. of California Press,

politics Berkeley.
Rakodi, C. (1995): Rental tenure in the Cities of Developing
Countries. Urban Studies, 32, 4-5: 791-811.

Strassman, W.P. (1985): Types of neighbourhoods and homebased
enterprises: evidence from Lima, Peru. Michigan State
University, East Lansing.

Strassman, W.P. (1987): Home-based enterprises in cities of
Developing Countries. Economic Development and Cultural
Change, 36, 1:121-

Tipple, A.G. (1993): Shelter as workplace: a review of homebased
enterprises in developing countries. International Labour
Review, 132,4:521-539.

UNCHS/ILO (1995) Shelter provision and employment generation.
Nairobi, UNCHS and Geneva, ILO.