Land Title Ambiguity and Underinvestment

1. The Question

Agriculture forms the livelihood for around 60% of the rural population in Kenya, with these people cultivating relatively small areas of land ranging from 2-5 acres; however, less than one in five people living in the rural areas of Kenya actually have their land officially titled. Development economics has always considered lack of legal titling to be a major reason why investment in rural lands is underdeveloped; since a person cannot obtain a loan against untitled land, he/she will make very limited investments because of the uncertainty regarding future use of land (for instance soil conservation, tree plantation, terracing).

Kenya provides a very strong test to this hypothesis as it has been practicing this form of land adjudication and registration since colonial times, and even immediately after the attainment of independence. The question addressed by this paper is: in the specific case of Kenyan smallholders, does the empirical record actually support the de Soto chain, title leads to credit leads to investment, or does it reveal a different, more specific constraint that titling alone does not resolve?

2. The Microeconomic Mechanism

The de Soto mechanism rests on a sequence of conditions, each of which must hold for titling to raise investment:

1. Enforceability condition: a title of a loan asset should provide the creditor enough assurance that it can be seized and sold in case of default. 2. Scale condition: The value of the loan should exceed that of the foreclosure and transaction costs by an adequate margin so that the lender makes profit from financing the loans to the farmers. 3. Uniqueness condition: the title should have the sole legal entitlement to the land, which means that there shouldn’t be any competing titles, such as customary rights, or inheritance rights. 4. Investment-response condition: when credit becomes available, farmers need profitable credit constraints waiting to be exploited.

Each of these poses an independent empirical question, and indeed the second and third have been difficult to satisfy in the context of Kenya’s land economy. Indeed, Kenyan smallholder farms are often subdivided over the generations through inheritance, resulting in plots of one or two acres in size or smaller, which, on account of their resale value being lower than the fixed costs involved in appraising them as loan collateral (such as the cost of registering a deed and verifying its status), may fail to be deemed commercially attractive enough by banks. In addition, despite registration, customary and familial claims to land often remained in practice, even where there was formal registration (as seen throughout the literature on land reform in East Africa): siblings, members of an extended family, or customarily recognized bodies might continue to make informal claims which were not extinguished by formal registration, particularly where registration took place decades ago and family dynamics had since changed through inheritance or informal subdivision.

3. Evidence

The classic Kenyan test of the de Soto hypothesis found weak effects. The Case of Migot-Adholla’s study, which is commonly referred to, of land titling in Nyeri and Kakamega districts tested for the effects of land titling in regard to tenure security, formal credit access, agricultural productivity, and land concentration. The study found out that land titling did not have a significant effect on credit access and conditions, productivity of crops, and concentration of land, and hence deduced that the farmers were mainly interested in titles as an indicator of security of rights as opposed to production of crops.

More recent Kenyan studies find a real, though more qualified, link. In one recent study, conditional mixed process estimation based on Kenyan smallholder data shows that land tenure security is positively correlated with access to credit and that credit access positively impacts agricultural productivity, providing support for the collateral theory, albeit insofar as this effect operates via credit access as a mediating variable rather than being an inherent effect of land titling itself. Another cost-benefit analysis of land tenure security schemes implemented in Sub-Saharan Africa found that land tenure security programs increased borrowing in Kenya by around 9 percentage points, with the credit accessed being used not only for agricultural purposes but also for investments in human capital, businesses, and health care, showing that any credit-unlocking effects are beneficial beyond simply increasing farm productivity.

The scale problem is directly visible in current financing-gap statistics. Modern-day reports on the state of agricultural finance in Kenya are unequivocal on the figure: roughly 80% of the country’s smallholder farmers and small-scale agribusinesses are unable to secure formal loans from banks precisely because they lack a formal land title that can be used as collateral, resulting in an agricultural funding gap of several billion dollars. When combined with the fact that only about one in every five rural residents in Kenya has a formal land title, it is clear that the problem is a binding one. However, that does not make a distinction between whether titles do or do not exist.

The market’s own behavior is the strongest evidence that “title” was never the whole story. If formal land title were a clear and adequate solution to the issue of collateral, then it would be reasonable to see an overwhelming focus of financial innovation on the expansion of titling. However, what we actually see in Kenya is that the most innovative solutions in agricultural finance do not rely on titling at all: mobile lending (Tala and others) underwrites loans based on digital transaction records of borrowers and manages to achieve a repayment rate of 89% for small agribusinesses without any need for formal land titling; legal changes made in 2019 (Warehouse Receipt System Act) allow the use of stored and certified commodities as collateral instead of land; and finally, legal changes made within the Movable Property Act allow farmers to put up livestock as collateral and achieve a year-on-year increase in pledges of livestock and financial securities by about 53.6% as of mid-2025. All this suggests a clear preference of the financial sector of Kenya: lenders and regulators are developing collateral substitutes for the land-titling system, not within the system.

4. Argument

The Kenyan data, viewed as a whole, actually do not disprove the de Soto hypothesis but rather define the true breaking point more accurately. First of all, the poor findings of the classic experiment in Nyeri and Kakamega cannot be used to conclude that collateral is irrelevant to access to credit because more recent research proves its relevance in the situation where credit access is treated as the mediating variable. Secondly, the poor findings of this experiment arise not from the fact that registration does not guarantee collateral, which is crucial for credit access but from the fact that registration per se is not enough to make it bankable due to the fact that the scale and enforcement problems still exist.

This reframing is critical because it shifts where the investment should be targeted. In the case where “farmers don’t have titles” was considered the problem, the solution prescribed would be the rapid registration of farmers. But when the problem is that even the titled land is too fragmented and encumbered by informal claims to work as a credible collateral security, the solutions will be found in the market innovations in Kenya, which bypass the land system completely: alternative forms of collateral (warehouse receipts, movable collateral, loan based on transaction history). That these alternative forms of collateral are developing rapidly and in the case of mobile lending, have very high repayment ratios indicates that the investment-response condition (condition 4 above) remains valid: there are profitable investments available for financing in Kenya’s rural areas. The problem was not in the lack of any collateral but in using the wrong form of collateral, namely, the formally titled land.

5. Implications

• For policy: However, the further pursuit of land registration as a mechanism to provide tenure security and the non-credit gains that accompany tenure security (land dispute reduction, clear inheritance rights, improved perceptions of security) is still worthwhile; it should not, however, be presented or financed as a credit access solution, the evidence from Kenya suggests otherwise.
• For finance-sector strategy: These instruments of collateral-substitution (i.e., warehouse receipts, movable asset registries, and alternative data lending), which seem to be addressing the real issues of scale and enforceability, should get more regulatory backing and expansion going forward.
• For future firm/household-level research: A valuable follow-up to the study by a group of student researchers could be an examination of the scale problem hypothesis by contrasting the access to credit and investment results for titles on plots larger than the critical size level against those on plots below it, along with the existence of informal claims against the titled plots.

6. Limitations

While this paper summarizes the available survey and literature evidence on the topic, not new household data, there are significant differences between the two major Kenyan studies cited in this paper (Place & Migot-Adholla and the more recent conditional mixed-process study), which prevent making any direct comparisons between them-the “weak effect” study was conducted decades ago in Nyeri and Kakamega regions, while the “positive effect” study is based on recent survey data. The scale problem and persistent informal claims mechanisms identified as the reason for this difference look quite plausible and consistent with general findings in the field of land economics, although they are not tested against Kenyan data in this paper-such an analysis would require plot-level data on title status, plot size, existence of competing informal claims, results of loan applications and investment behavior. Finally, financing gap and repayment rate numbers provided in this paper have been cited from non-peer-reviewed industry and advocacy sources, therefore should not be regarded as exact.

References (selected)

1. Place, F. & Migot-Adholla, S. E. (1998). The economic effects of land registration on smallholder farms in Kenya: evidence from Nyeri and Kakamega districts. Land Economics, 74(3).
2. Mbudyza, J. J. et al. (2022). Effect of land tenure security on agricultural productivity among small-scale farmers in Kenya: a conditional mixed processes analysis. Cogent Economics & Finance.
3. Aberra, D. & Chemin, M. (2021). (Cited in the Journal of Benefit-Cost Analysis cost-benefit review of land tenure security in Sub-Saharan Africa, including Kenya credit-union borrowing effects.)
4. de Soto, H. (2000). The Mystery of Capital: Why Capitalism Triumphs in the West and Fails Everywhere Else.
5. Ministry of Lands and Physical Planning, Kenya (2024), rural land titling coverage statistics.
6. African Development Bank (2024), agricultural financing gap estimates.
7. Kenya’s Warehouse Receipt System Act (2019) and Movable Property Security Rights Act, legal frameworks for alternative collateral.
8. Business Registration Services (BRS), Kenya, movable-asset pledge growth data.
9. USAID Kenya Investment Mechanism / Agrilinks, reporting on credit guarantees and agricultural finance innovation.

Prepared as part of the East Africa Microeconomics Research Series. This is a student working paper that synthesises existing research and published data rather than presenting original fieldwork; feedback and data corrections are welcome via the contact page.