Helle Munk Ravnborg
Danish Institute for International Studies, Copenhagen, Denmark
College of Agricultural and Environmental Sciences, Makerere University, Uganda
International Livestock Research Institute, Nairobi, Kenya
Rasmus Hundsbæk Pedersen
Danish Institute of International Studies, Copenhagen, Denmark
Department for Society and Globalisation, Roskilde University, Denmark
College of Agricultural and Environmental Sciences, Makerere University, Uganda
This article on the results of a questionnaire survey conducted in the areas of Amuru (comprising Amuru and Nwoya districts), Masaka (comprising Bukomansimbi, Lwengo, Kalungu and Masaka districts) and Pallisa (Butebo and Pallisa districts) in mid-2012 forms part of a study of the linkages between land and property rights and economic behaviour in Uganda. The study was commissioned by the Royal Danish Embassy in Kampala and carried out by a team of researchers from the Danish Institute for International Studies, Copenhagen (DIIS) and Makerere University, Kampala (MUK).
The questionnaire survey was completed by a total of 1,174 respondents in the three areas. Only respondents currently holding access to land were sampled. The questionnaire was prepared on the basis of exploratory, qualitative interviews conducted in the three areas into issues related to land tenure, security of tenure and economic behaviour; and on the basis of a review of the literature on these issues, primarily in Uganda (Pedersen et al., 2012). In addition to collecting information on land tenure and economic activities, the questionnaire was designed to provide the data necessary to develop a poverty profile of the households to which the respondents belonged, distinguishing between non-poor households, less poor households and the poorest households. For more detail, see Ravnborg et al. (2013).
Tenure security is widespread – and mostly so in Amuru
Contrary to the picture often given (e.g. Rugadya 2008), the majority of people who currently have access to land in the Amuru area in northern Uganda feel confident that they will continue to be able to access the land they are currently using. Specifically, security of tenure is more widely perceived among respondents in Amuru than among respondents in Masaka and Pallisa. As part of the questionnaire survey a total of 1,174 respondents in the three areas were asked to assess ‘how secure they feel in their land tenure’ with respect to a total of up to three parcels of land that they currently access. As a result, this assessment was made with respect to a total of 2,163 land parcels.
In the Amuru area, respondents perceive themselves to be ‘very secure’ in their land tenure with respect to 40 per cent of the parcels and ‘secure’ with respect to a further 52 per cent of the parcels (Table 1). Although land tenure is widely perceived to be secure in Pallisa, almost a quarter of all respondents in Pallisa (22 per cent) consider their tenure to be ‘not that secure’ (Table 1), meaning that tenure is perceived to be insecure significantly more widely in Pallisa than in Amuru and Masaka.
Table 1: Perceived Security of Tenure, Amuru, Masaka and Pallisa Areas (N=2,163 parcels)a
Percentage of parcels by access holders’ overall perceived level of security of tenure
a Significant correlation between distribution of parcels according to perceived security of tenure and area at 0.001 level (Pearson Chi-Square).
Tenure security may mean different things to different people living in different cultural and economic settings. Informed by the literature, we explored the extent to which the following three points were considered by respondents when assessing their tenure security:
- the ability to uphold rights of continued use of the land;
- the ability to reap benefits of invested labour and capital by bequeathing land to children or close relatives;
- the ability to reap benefits of invested labour and capital by selling the land.
Respondents were asked to assess their security of tenure on these three points for each of the land parcels covered by the survey. For all three areas – Amuru, Masaka and Pallisa – the results show a close positive correlation between the perceived level of tenure security (shown in Table 1 above) and the perceived level of security with respect to points 1 and 2 (continued use of the land and the ability to bequeath it to children or close relatives), whereas such a positive correlation does not exist with respect to point 3 (their ability to sell the land). First of all, respondents expect to be able to sell less than three per cent of the land parcels. For the vast majority of the parcels (90 per cent), respondents are unsure of their ability to sell; and for the remaining eight per cent of parcels, respondents expect to be able to sell, but only with the approval of others and without knowing exactly who those others might be.
As well showing that the ability to sell land is not seen as an integral and fundamental part of land tenure security in the three areas, these findings also provide a more general indication of the current state of the land market in Amuru, Masaka and Pallisa. We shall return to this point.
Tenure insecurity is associated with the co-existence of tenure forms rather than with any specific tenure form
In the literature, it is common to associate perceived tenure security with, on the one hand, the particular form of tenure under which land is held, and, on the other hand, with the level of documentation and registration of land rights, e.g. formally registered land titles (e.g. Deininger 2003; Feder & Feeny 1991; de Soto 2000). In Uganda, five forms of land tenure are recognised, namely (i) customary tenure; (ii) mailo tenure and the related (iii) kibanja tenancy; (iv) freehold tenure; and (v) leasehold. In the context of this survey, mailo and the associated kibanja tenure are only found in the Masaka area (Table 2), while customary tenure is the most widespread tenure form in Amuru and Pallisa. In the Amuru area, hardly any land is held under other tenure forms, including freehold, whereas around 15 per cent of the parcels in Masaka and Pallisa are held under freehold tenure.
Table 2: Forms of Land Tenure, Amuru, Masaka and Pallisa Areas (N=1,172 respondents; information missing for 2 respondents; N=2,218 parcels; information is missing for 53 parcels)a
Percentage of respondentsb and percentage of parcels per area by tenure form
|Area||All tenure forms|
|Amuru (n=398 respondents||95.7||–||2.8||–||–||6.6||–b|
|Masaka (n=384 respondents)||14.6||62.2||16.9||8.1||2.3||9.9||–b|
|Pallisa (n=390 respondents)||89.7||–||23.6||–||4.9||7.0||–b|
|All areas (N=1,172 respondents)||67.2||20.4||14.3||2.6||2.4||7.8||–b|
|Amuru (n=895 parcels)||93.7||–||1.8||–||–||4.5||100.0|
|Masaka (n=623 parcels)||10.8||56.3||14.6||7.2||1.8||9.3||100.0|
|Pallisa (n=100 parcels)||75.1||–||16.3||–||3.6||5.0||100.0|
|All areas (N=2,218 parcels)||64.6||15.8||10.0||2.0||1.6||6.0||100.0|
a Significant correlation between each of the tenure forms and area both for respondents and parcels at 0.001 level (Pearson Chi-Square).
b Respondents who have provided information about more than one parcel may hold their parcels under different tenure forms. Therefore respondent percentages add up to more than 100.0.
In Masaka, parcels held under freehold and mailo tenure are associated with a higher level of perceived tenure security than parcels held under kibanja tenancy. Thus, while rights-holders to 70 per cent of the parcels held under mailo tenure and to 60 per cent of the parcels held under freehold tenure assess their land tenure to be ‘secure’, this is only true for the rights-holders to 46 per cent of the parcels held under kibanja tenure. In Pallisa, respondents holding parcels under freehold tenure are more likely to assess their land tenure as ‘secure’ than respondents holding land under customary and particularly under leasehold tenure. In Amuru, however, where the assessment of land tenure as ‘secure’ is most widespread, applying to 62 per cent of the parcels, and where virtually all land is held under customary tenure, no correlation is found between the form of land tenure and the level of tenure security.
This suggests that rather than being associated with the tenure form itself, perceived tenure security – or rather insecurity – may be associated with the co-existence of different forms of tenure. This stems from the fact that the co-existence of different forms of tenure often implies ambiguities over the values and standards according to which claims are made, the co-existence of – and often competition between – different authorities through which claims are made, and, more profoundly, that the co-existence of different forms of land tenure may result from more profound societal transformation, in the course of which supremacy may eventually be assigned to one tenure form and the institutions associated with it.
Mode of land acquisition constitutes an important additional aspect of land tenure and acquiring land through rental agreements is associated with tenure insecurity
Both during the exploratory interviews conducted in the three areas prior to the questionnaire survey and during the questionnaire survey itself, many respondents mentioned the mode of acquisition of a particular parcel in the same breath as naming the tenure form under which the parcel was held, thereby adding an additional dimension to the tenure form.
The three areas vary significantly with respect to the prominence of different land acquisition forms (Table 3). In Amuru, virtually all parcels (88 per cent) were inherited, whereas in Masaka the majority (66 per cent) of the parcels were purchased and only a quarter (25 per cent) was inherited. Pallisa falls between the two other areas in this respect. Although the majority of the parcels were inherited, 17 per cent were purchased and an additional eight per cent accessed through rental agreements.
Table 3: Land Acquisition Mode, Amuru, Masaka and Pallisa Areas (N=2,255; information is missing for 16 parcels)a
Percentage of parcels per area by land acquisition mode
a Significant correlation between land acquisition mode and area at 0.001 level (Pearson Chi-Square).
In all three areas, the level of perceived tenure security is correlated with the way in which land was acquired. In Masaka and Pallisa, where buying of land is most common, respondents are more likely to assess their tenure to be ‘secure’ for parcels that have been purchased than for parcels acquired in other ways. At the other end of the security scale, parcels that are rented are the most likely to give rise to the perception of tenure insecurity. Respondents perceive their tenure to be ‘not that secure’ in the case of 69 per cent of the parcels that are rented compared to six per cent of all parcels acquired in other ways.
There is no empirical justification for equating tenure security with holding formal land titles. Although having precise written documentation of land tenure reduces tenure insecurity, having no written documentation of land tenure by no means excludes tenure security
In the land tenure literature, holding a formal land title is often equated with tenure security (e.g. Deninger 2003; Feder & Feeny 1991). The widespread perception of tenure security in Amuru, Masaka and Pallisa combined with the limited proportion of parcels of land for which formal titles are held clearly shows that this equation is not empirically justified. Formal titles are held for less than one fifth of the parcels (Table 4); yet the respondents assess their land tenure to be ‘secure’ in the case of more than half of the parcels.
Table 4: Land Title held, Amuru, Masaka and Pallisa Areas (N=2,264 parcels; information missing for 7 parcels)
Percentage of parcels per area by title held
|Area||Type of title held|
|Certificate of customary tenure||Freehold title||Mailo title||Leasehold title|
|Amuru (n=917 parcels)||1.4||0.4||–||–|
|Masaka (n=638 parcels)||1.3||6.1||3.4||1.4|
|Pallisa (n=709 parcels)||27.9||8.6||–||0.6|
|All areas (N=2,264 parcels)||9.7||4.6||1.0||0.6|
However, in addition to formal titles a range of other written documents exists; these serve in different ways to document land tenure. They include what may be labelled (i) private documentation, i.e. documentation between private individuals such as a will, rental agreement, sketch map, purchase agreement, etc.; (ii) partial formal documentation, i.e. documentation involving or relying upon third party authorities such as receipts of paid property tax, proof of application for certificate of customary tenure, freehold tenure, etc.; and (iii) complete formal documentation such as certificate of customary tenure, freehold, leasehold or mailo title. In addition to the type of written documentation, a number of other characteristics of such documentation are also important in assessing its strength as tenure documentation for the respondent. These characteristics include whether the document is written in the name of the respondent or somebody else, whether it is registered with a third party or relevant authority (customary or statutory), and whether it describes the location and extension of the parcel.
Based on the type of written documentation and the above characteristics, a tenure documentation index was developed. According to the existence of written tenure documentation, its type and other characteristics, the parcels were divided into three categories, namely (i) those for which no written tenure documentation exists; (ii) those for which only partial or imprecise written documentation exists, e.g. a tax receipt or a will in the name of a relative; and (iii) those for which more complete and more precise written documentation exists. Table 5 shows the distribution of parcels according to these tenure documentation categories for the three areas.
The extent to which tenure is documented varies significantly between the three areas. Hardly anybody in Amuru has written documentation of their land tenure – written documentation only exists for five per cent of the parcels in Amuru – while written documentation exists for more than 80 per cent of the parcels in Masaka, with only private documentation for nearly 60 per cent of the parcels. In Pallisa, no written documentation exists for 40 per cent of the parcels. However, for the rest of the parcels, fairly complete documentation exists for the majority.
In all three areas, having written and precise tenure documentation is associated with a reduced likelihood of perceiving tenure to be ‘not that secure’. But the evidence is mixed with respect to having written and precise tenure documentation being associated with a higher likelihood of perceiving tenure to be ‘secure’. This is the case in Masaka but not in Pallisa.
Although respondents in Masaka who do not hold any written tenure documentation are less likely to perceive their land tenure to be ‘secure’ than respondents who hold some or more precise documentation, no such correlation is found in Pallisa.
As already mentioned, in Amuru written tenure documentation only exists for five per cent of the parcels. Nevertheless, respondents assess their tenure as ‘secure’ with respect to 59 per cent of the parcels. Put differently, having no written documentation to support their claims to a right to access a particular piece of land by no means excludes respondents from perceiving their land tenure to be ‘secure’.
As an additional indication of the different importance assigned to written tenure documentation in the three areas, respondents have plans to improve the documentation of their tenure in the case of half (51 per cent) of the parcels in Masaka while this was the case for only 15 per cent of parcels in Pallisa and just 6 per cent in Amuru. In the overwhelming majority of cases in Masaka as well as in Pallisa and Amuru, the improvement wished for was to obtain land titles.
Land investments do not depend on land security
A large proportion of respondents in Masaka and Pallisa make agricultural investments in their land. The majority of this investment is primarily labour intensive whereas capital intensive investment, such as the establishment of irrigation, is less common. Following the peace agreement making it possible for people in northern Uganda to return to their land, people in Amuru have been in the process of ‘settling in’, i.e. rebuilding their houses and opening up their land. This probably helps to explain why, over the last five years, agricultural investment (apart from opening up land) has been sparse in Amuru.
Investment is equally likely in parcels for which tenure is perceived as ‘secure’ as in parcels for which tenure is perceived as ‘somewhat secure’ or ‘not that secure’ (Table 5) as well as in parcels held under different tenure forms (Table 6).
Table 5: Agricultural Investment by Level of Perceived Tenure Security, Amuru, Masaka and Pallisa Areas (N=2,123 parcels; information missing for 148 respondents)a
Percent of parcels where agricultural investment has been undertaken per tenure security level, by area
|Area||Level of perceived tenure security||All tenure security levels|
|Secure||Somewhat secure||Not that secure|
|Amuru (n=869 parcels)||1.7||0.3||0.0||1.2|
|Masaka (n=622 parcels)||50.6||55.0||48.6||52.4|
|Pallisa (n=632 parcels)b||57.9||60.6||44.0||57.9|
|All areas (N=2,123 parcels)||30.3||37.0||32.5||33.1|
a Parcels that are rented and thus from the outset associated with restrictions on investment are omitted from this cross-tabulation.
bp=0.093; Pearson Chi-Square test.
Table 6: Agricultural Investment by Tenure Form, Amuru, Masaka and Pallisa Areas (N=2,032 parcels; information missing for 239 respondents)a
Percentage of parcels where agricultural investment has been undertaken per tenure form, by area
|Area||Tenure form||All tenure forms|
|Amuru (n=849 parcels)||1.2||–||–||–||–||1.2|
|Masaka (n=557 parcels)||42.4||54.4||53.8||57.8||36.4||52.8|
|Pallisa (n=651 parcels)b||59.7||–||51.4||–||12.0||56.5|
|All areas (N=2,057 parcels)||24.5||54.4||48.6||57.8||19.4||32.7|
a Parcels that are rented through private land rental agreements are omitted from this cross-tabulation.
b Significant correlation between tenure form and having undertaken agricultural investments at 0.001 level (Pearson Chi-Square).
Absence of land titles is not a main factor limiting demand for formal credit among respondents who currently have access to land
Despite concerted efforts during the past decade or two to increase the supply of agricultural credit e.g. by providing subsidized capital funds to credit institutions, demand has not increased in practice as anticipated. Overall, only around a quarter of respondents had taken out a loan and the vast majority of those who had taken loans during the past five years had done so to finance non-productive investments, such as education for children, meeting health expenditures, repairing their houses, etc. (Table 7).
Table 7: Taking Loans to Finance Productive and Non-productive Investments, Amuru, Masaka and Pallisa Areasa (N=1,174 respondents)
Percentage of respondents taking loans by type of investment financed, by area
|Area||No loan taken||Type of investment financed through loana||All types of investments financed through loan|
|Agricultural investments, only||Non-agricultural productive investments, only||Non-productive investments, only||Productive & non-productive investments|
|Amuru (n=399 respondents)||94.5||0.3||0.3||4.5||0.5||5.5|
|Masaka (n=384 respondents)||72.1||1.3||0.3||20.8||5.5||27.9|
|Pallisa (n=391 respondents)||68.0||0.8||0.0||29.4||1.8||32.0|
|All areas (N=1,174 respondents)||78.4||0.8||0.2||18.1||2.6||21.6|
a Significant correlation between area and taking loans to finance productive and non-productive investments at 0.001 level (Pearson Chi-Square).
Of those who do take loans, rather less than half take formal credit, sometimes combined with informal loans, whereas the rest only take informal credit (Table 8).
Table 8: Loan Taking, Amuru, Masaka and Pallisa Areasa (N=1,174 respondents)
Percentage of respondents by loan taken, by area
|No loans taken||Loan taken through formalb credit||Loan taken through informalc credit||All respondent|
|Amuru (n=399 respondents)||94.2||2.5||3.3||100.0|
|Masaka (n=384 respondents)||71.9||15.4||12.8||100.0|
|Pallisa (n=391 respondents)||67.8||10.0||22.3||100.0|
|All areas (N=1,174 respondents)||78.1||9.2||12.7||100.0|
a Significant correlation between area and type of loan taken at 0.001 level (Pearson Chi-Square).
b SACCOs (Savings and Credit Cooperative Societies) and banks.
c Saving groups, private individuals, etc.
Only around one third of respondents taking formal credit had used land as collateral (Table 9).
Table 9: Collateral Used for Formal Loan Taking, Amuru, Masaka and Pallisa Areasa (N=107 respondents)
Percentage of respondents by type of collateral used, by area
|No collateral used||Land used as collateral||Other type of collateral used||All respondents taking formal credit|
|Amuru (n=10 respondents)||20.0||40.0||40.0||100.0|
|Masaka (n=58 respondents)||22.4||44.8||32.8||100.0|
|Pallisa (n=39 respondents)||17.9||17.9||64.1||100.0|
|All areas (N=107 respondents)||20.6||34.6||44.9||100.0|
a Significant correlation between area and type of collateral used at 0.05 level (Pearson Chi-Square).
A bigger share of those who hold land titles use their land as collateral for their loans than of those who do not hold titles (34 per cent as compared to 18 per cent). However, the majority – also of respondents holding titles in support of their land tenure – use assets other than land as collateral for their loans or take loans that do not require collateral. Thus, while the absence of a formal title by no means excludes respondents from taking loans, respondents who hold land titles are more likely to present collateral – both land and other assets – when taking loans than respondents who do not hold land titles. This indicates that holding land titles may also correlate with other factors, such as having non-agricultural sources of income which facilitate access to credit requiring the use of collateral.
Tenure insecurity is associated with the co-existence of different tenure forms rather than with any tenure form in particular
No particular tenure form has the monopoly on providing tenure security. In Amuru, almost all land is held under customary tenure and at the same time, compared to both Masaka and Pallisa, it is the area where tenure is perceived as ‘secure’ with respect to the highest proportion of parcels. By contrast, in Pallisa where freehold tenure has become more widespread during recent decades, holding land under customary tenure is associated with a lower likelihood of perceiving tenure as ‘secure’ as compared to holding land under freehold tenure. Rather than being an absolute feature derived from any particular tenure form, the level of perceived tenure security is relative in the sense that it depends upon the extent to which other people’s tenure rights, of whatever form, are protected. As soon as new tenure forms and new forms of written tenure documentation emerge, this affects the perception of tenure security of land tenure holders at large, particularly if preferential treatment is given to specific forms of tenure and tenure documentation.
Written tenure documentation is appreciated, but tenure may be perceived as secure without it
As in most other places, it is hard to find anyone who, if given the choice, would opt not to have a title deed endorsing his or her tenure rights to a piece of land. Yet, the pattern of correlation is far from straight-forward and uniform between, on the one hand, holding a title or any other form of written tenure documentation and having it registered with relevant authorities (customary as well as statutory) and, on the other hand, the perception of tenure security. While holding no, or only incomplete, tenure documentation at least in Masaka and Pallisa is associated with tenure insecurity, it is only in Masaka that holding more complete written tenure documentation convincingly translates into an increased likelihood of tenure being perceived as ‘secure’.
Context and the resource endowments of the individual are important constituents of tenure security
Rather than suggesting that written tenure documentation does not bolster the perception of tenure security, these findings indicate that titles and other written tenure documentation are only one among several elements that together create the perception of tenure security. Among these additional elements are both context- or area-specific and respondent-specific features. The area-specific features include the rules and norms guiding the institutions backing different forms of land tenure and the relative strength of these institutions, while the respondent-specific features include the location of residence, household poverty level, ethnicity and sex of the respondent, as these characteristics contribute to determine the ease with which land can be held under different tenure forms and the ease with which the institutions backing a particular land claim can be mobilised.
Socio-economic inequality translates into inequality in perceived tenure security
In Amuru, where people have only recently resettled and where the notion that ‘you can only claim as much land as you can dig’ may still be encountered, land distribution is still relatively equal, compared to the other two study areas. The distinction between customary and statutory institutions is blurred and often, in Amuru as elsewhere, those who are recognised as clan leaders are also likely to hold offices with parish and sub-county level statutory institutions. As virtually all land in Amuru is held under customary tenure without the support of written documentation, community and clan membership and relations are important in supporting land claims. Thus, in contrast to Masaka and Pallisa, respondents in Amuru who belong to the poorest households and reside in rural areas are equally – if not more – likely to perceive their land tenure to be ‘secure’ as their less poor or peri-urban neighbours, while the few non-Acholi respondents holding land in the area seem less likely than the Acholi respondents to perceive their tenure as ‘secure’. Although access to land under customary tenure as practised in Amuru is far from equal for men and women, female and male respondents are equally likely to perceive their tenure to be ‘secure’.
The fact that hardly anybody in Amuru holds written documentation in support of their tenure renders this point insignificant to most people in the area when assessing their tenure security. However, those who do hold fairly complete written tenure documentation are more likely to perceive their tenure as ‘secure’. This, however, may change. As more people apply for and obtain certificates of customary tenure, for example, and as new tenure forms emerge, the role of written tenure documentation in shaping tenure security perceptions will change, even among those who do not hold such documentation.
Land disputes will always exist and broad-based access to institutions is important in preventing land disputes from translating into generalised tenure insecurity
In addition to illustrating the relative – rather than the absolute – importance of tenure form and written tenure documentation, the results from Amuru also serve to illustrate the importance of access to institutions which can mediate directly in cases of land disputes and, in cases where disputes and other tenure-related issues cannot be resolved locally, can facilitate access to higher-level institutions. Due to the massive resettlement which has taken place since 2006, Amuru is the area with the highest incidence of land disputes among the respondents, while at the same time being the area where the perception of tenure security is most widespread. The strong engagement by non-governmental organisations in informing people about land laws and land rights and facilitating the establishment and training of local-level dispute settlement mechanisms may have contributed to this. Combined with testimonies from interviewees, this points to the importance of institutions that are accessible to the broad majority rather than only to a small segment of the population.
Leaving aside other problems, this situation, at least from the perspective of tenure security, appears both ideal and inclusive. But it may be temporary – like the silence both before and after the storm. So, action is called for, just as it is in other parts of Uganda. If the aim is to maintain and deepen current high levels of perceived tenure security, while at the same time unlocking the land market, important lessons for policy and administrative interventions may be drawn from observing the situation in Masaka and Pallisa.
Land markets as a double-edged sword, reducing land access for some while facilitating land access for others
Masaka, where the majority of land transactions, irrespective of tenure form, are mediated through the market and where customary tenure, whether labelled as such or as kibanja tenancy, is undergoing profound transformation, is also the area where belonging to a non-poor household and residing in an urban or peri-urban area most clearly translate into increased security of tenure. Compared to both Amuru and Pallisa, Masaka is also the area where customary institutions are least solicited and where respondents holding land under customary tenure appear to have the most limited choice of institutions from which to seek support if their land rights are challenged, whereas respondents holding their land under freehold tenure in such cases appear to have a much more diverse set of institutions to call upon in their defence.
In Pallisa, there is a correlation between, on the one hand, household poverty level and residence and, on the other hand, tenure security, but it is strongly confounded by the correlation between the sex of the respondent and tenure security. The norms embedded in customary institutions in Pallisa first of all limit women’s access to land and, secondly, imply that women who do succeed in gaining access to land are much more likely to perceive their tenure as less secure than men in the area. Such differences between male and female respondents are not found in Amuru and Masaka. Although the majority of the female respondents who access land in Pallisa do so under customary tenure, a significantly higher proportion of female than male respondents have recourse to land rentals, which under current conditions are perceived to be insecure and associated with significant limitations in terms of investment. A similar tendency is observed for non-Acholi respondents in Amuru who, presumably because they are not able to obtain access to land through customary institutions, are also more likely than their Acholi neighbours to rent land. Thus, land markets, including rental markets, have the potential to provide access to land for those who due to non-economic factors, such as gender or ethnicity, are otherwise excluded from gaining access, while at the same time tending to favour the economically resourceful.
Credit in its present forms does not limit productive investment
There is only limited evidence that productive agricultural and non-agricultural investment is hampered by low levels of tenure security, absence of formally registered tenure documentation and lack of access to credit in its present forms. The absence of the often assumed relationship between perceived tenure security and investment suggests that the inverse relationship also exists, namely that investment is undertaken as part of a strategy of strengthening land claims and thus tenure security, adopted by kibanja tenants in Masaka, for example.
Although nearly all the respondents interviewed in Amuru since 2006 have returned to their place of origin and have spent the past five to six years (re)building their homes and opening up their land, very few of them reported having made any investments in the last five years. By contrast, in Pallisa and Masaka, the majority of respondents (two-thirds and three-quarters, respectively) have made some form of productive investment in the past five years. However, the vast majority of these investments were financed through their own labour and savings, rather than through credit.
Credit is used to finance non-productive rather than productive investments
Nevertheless, credit is used – in Masaka by around a quarter of respondents and in Pallisa by a bit less than a third of respondents. Rather than being used to finance productive investments, it is used to finance non-productive investments like education for children, health expenditure, etc. While by no means being exclusively taken by non-poor people, respondents belonging to non-poor households are more likely to take loans than those belonging to poorer households.
Land is neither the preferred collateral nor is it required as collateral
Land was only used as collateral for around one-third of the loans taken with formal credit institutions such as Savings and Credit Cooperatives (SACCOs) and banks and only in one fifth of these cases (n=7) was land tenure supported by a formal land title. Credit institutions appear to accommodate a certain degree of flexibility with respect to land tenure documentation to the extent that local institutions – statutory as well as customary – are approached to provide their endorsement of the tenure rights of the loan applicant and their consent to the loan agreement. Rather than indicating restrictions imposed by credit institutions, the limited use of land as collateral appears to reflect a widespread hesitation by the population in this regard. Beyond the low level of trust in credit institutions and the legal system, this may be due to the widespread desire not to put at risk something (land) which is regarded as not only belonging to the individual but to the family, the clan or the community at large (the perception of not having the right to sell land). Instead of land, other and more individual assets are used as collateral, such as salary accounts, and this tends to favour the access to credit of non-poor individuals.
Tenure security is currently widespread among land holders in the three study areas. Overall, tenure is perceived to be ‘secure’ with respect to half of the parcels included in the survey while being perceived as ‘not that secure’ with respect to less than 10 per cent, leaving the tenure of the remaining 40 per cent (approximately) of the parcels perceived to be ‘somewhat secure’. However, the fact that tenure is currently perceived to be secure does not preclude the possibility that land tenure may be lost, for example to outside investors. The numerous press reports of land grabbing and land conflicts, not least in the northern part of Uganda, indicate that such risks may be real.
Uganda is currently in the process of reforming its National Land Policy. Although national land policies, as well as the legislative and administrative frameworks necessary for their implementation, should not be expected to be static, our results suggest that their very revision may nurture a growing perception of tenure insecurity. Tenure insecurity is found to be associated with the co-existence of different tenure forms rather than with any tenure form in particular.
Co-existing tenure forms, and the perceived level of tenure security they contribute to induce, are best seen as a series of mutually communicating vessels: If, for example, only institutions backing individual land claims are supported or if the institutions supported are accessible only to a small segment of the population, such partial efforts will contribute to generate tenure insecurity among those who wish to maintain their collective land rights or who are unable to access the institutions for receiving support.
If the aim is to maintain, deepen and widen the perception of tenure security among land holders in Uganda, and where possible, expand this perception to groups who today in some areas experience tenure insecurity, land administration interventions that are partial in scope or in coverage should be avoided as they are prone to capture by the elite and tend to induce, rather than reduce, tenure insecurity.
de Soto, H. (2000). The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else. New York: Basic Books.
Feder, G. & D. Feeny (1991). ‘Land Tenure and Property Rights: Theory and Implications for Development Policy,’ The World Bank Economic Review 5(1): 135–53.
Pedersen, R. H., Spichiger, R., Alobo, S. & Kidoido, M. (2012). ‘Land Tenure and Economic Activities in Uganda: A Literature Review.’ DIIS Working Paper 2012: 13. Copenhagen: Danish Institute for International Studies.
Ravnborg, H.M., Bashaasha, B., Pedersen, R.H., Spichiger, R. & Turinawe, A. (2013). ‘Land Tenure under Transition – Tenure Security, Land Institutions and Economic Activity in Uganda.’ DIIS Working Paper 2013:03. Copenhagen: Danish Institute for International Studies.
Rugadya, M.A. (2008). Northern Uganda Land Study. Analysis for Post Conflict Land Policy and Land Administration: A Survey of IDP Return and Resettlement Issues and Lesson from Acholi and Lango Regions. Kampala: World Bank.
 As part of the questionnaire survey, 399 respondents were interviewed in the Amuru area (Amuru and Nwoya districts) of whom 45 per cent were women. Virtually all of the respondents (95 per cent) had been displaced from the area during the past decades, the majority (62 per cent) for a period of more than 10 years. By only sampling and interviewing people who currently hold access to land, people who have not yet managed to return to their place of origin, some of whom are still living in refugee camps, and have not yet been allocated land, are not included in the survey.
 The mailo system was introduced by the colonial authorities in mutual agreement with the Buganda Kingdom in 1900. It gave the King and the feudal landlords freehold rights over large tracts of land, often inhabited by poorer subjects, who then became tenants of kibanja. This type of tenure system is prevalent in some regions of Uganda, for example Buganda, Bunyoro, Toro, Ankole and Bugisu. The mailo landowners and the Baganda leaders have opposed the national government’s efforts to gain control over land administration. (Ravnborg et al. 2013: 4).