Article of the Month
January 2008

Developing Cost-effective and Resilient Land Administration Systems in Latin America

Grenville Barnes; USA

This article in .pdf-format.

SUMMARY

In this paper I briefly review the wealth of experience in Latin America with initiatives to strengthen and modernize land administration systems. The review shows that there is more experience with land administration projects in this region than in any other. I go on to focus on the question of costs associated with formalizing property in an attempt to find an effective means of comparing costs across countries. I approach this by looking at different ‘levels’ – starting with global budget figures, then narrowing down to specific components and finally by examining the cost of individual tasks required to formalize a parcel. A comparative study of costs in 4 different regions of the world is described together with preliminary conclusions at the global level. The issue of property ‘deformalization’ is also discussed with respect to transaction costs.

Recognizing that land administration systems, and the economic, social and natural environments within which they operate, are continually changing, I introduce Resilience as an appropriate analytical framework through which to examine changing systems.
Resilience has evolved as a more nuanced framework for understanding the sustainability of socio-ecological systems. Unlike previous approaches, it accepts that a system will always be subject to disturbances, whether they are due to climate (hurricanes), policy and political administration changes, or demographic shifts due to urbanization or migrant labor markets.

RESUMEN

Esta exposición revisa la gran experiencia en América Latina sobre las iniciativas para fortalecer y modernizar los sistemas de administración de tierras. El repaso indica que hay mas experiencia en este region con pryetos de adminsitracion de tierras que en cualquier otra region del mundo. El primer parte enfoque en la cuestion de costos vinculado a la formalización de la propiedad con el motivo de identificar medios aptos para comparar costos entre diferentes paises. Esta analisis incorpora el estudio de costos en diferentes niveles – empezando con costos presupuestos globales, despues enfocando el nivel de componentes individuales y finalmente se examina los costos para diferentes actividades en el proceso de formaliza una propiedad. Se discutir un estudio de costos en cuatro diferentes regiones del mundo y ciertos conclusiones preliminares al nivel global. Esta discusión incluya tambien la cuestion de ‘deformalización’ de propiedad con respeto a los costos transaccionales.

Reconozco que sistemas de administración de tierras, y el ambiente económico, social e ecológico en que el sistema opera, siempre esta cambiando, se introduzca el Resiliencia como una rama analítica apropiada para analizar los dinámicos del sistema.

Finalmente, examinaré los sistemas de administración de tierras a través del lente de ‘resilencia.’ La resilencia se ha desarrollado como un marco matiz para examinar la sostenabilidad de sistemas socio-ecologico. A diferencia de previos acercamientos, éste acepta que el sistema siempre esté sujeto a disturbios, sea a consecuencia del clima (huracanes), cambios políticos o administrativos, o a desplazamientos demográficos a causa de la urbanización o migración de mercados laborales.

1. INTRODUCTION

It was appropriate to hold the recent regional FIG conference in Costa Rica as this country was one of the first countries in the region to implement what today would be regarded as a land administration project. The 1964 USAID-funded “cadastral survey project” was the pioneer land project in Central America (Goldstein 1974). The focus of that project was to improve the property tax system and complete a topographical mapping project which would provide information for tax, land planning and development purposes. Similar projects followed soon afterwards in Panama, Nicaragua and Guatemala, all focused strongly on property taxes, with additional components addressing natural resource management and land titling in some cases.

During the 1980s the World Bank began to fund several large land projects in Latin America and elsewhere. The first two to be completed were the projects in Thailand and NE Brazil (Holstein 1993). The Thailand “land titling project,” the ‘mother’ of all land titling projects, started in 1985. This has been a significant project for two reasons. Firstly, it is regarded as a highly successful project. Secondly, it has been the proving grounds for the evolutionary theory of land rights (ETLR) which served as the underlying rationale for many of the subsequent land projects that followed in the next two decades. Many of the assertions or hypotheses internal to the ETLR have been empirically proven using the experience of Thailand (Feder et al 1988). Data was gathered and analyzed to demonstrate the correlation between titling and access to credit, reduction in property disputes, facilitation of the land market, and increasing land values (Feder and Nishio 1996). Based partly on these positive outcomes in Thailand, land administration projects proliferated throughout the developing world.

Within Latin American the North East Brazil “national land administration project” was the forerunner of a series of land administration projects in the region (World Bank 1985). USAID continued to fund land titling projects throughout the 1980s, including the “land titling projects” in Honduras and Ecuador (USAID 1985). However, the most successful of the USAID-funded projects may have been the “Land Titling and Registration Project” in St. Lucia (USAID 1983). The island-wide land adjudication process was completed within the originally scheduled time – this by itself may be a unique achievement amongst land administration projects which invariably stretch beyond the time frame set out in the project design.

The World Bank has funded land administration projects throughout Latin America, and even by 1998 a study of World Bank projects revealed that there were “..115 projects with land-related activities in the Bank’s portfolio…[and] ..of those, about 40% are in Latin America.” (World Bank 1998, p. 10). Subsequent land administration projects ensued in Bolivia (1995, 2001), Brazil (1995), Guatemala (1996, 1997), Honduras (2000), Panama (2000), Nicaragua (2002), Honduras (2003), and El Salvador (2005).

The Inter-American Development Bank has also played a lead role in funding land administration projects in Latin America and the Caribbean especially over the past decade – including Trinidad and Tobago (1995), Nicaragua (1995), Dominican Republic (1997), Belize (1997), Colombia (1997), Honduras (1998), Jamaica (1999), Costa Rica (2000), Ecuador (2001), Panama (2002), Brazil (2002), Mexico (2003), Bolivia (2003), Paraguay (2003), and the Bahamas (2004).

While this is not a complete list of projects it does illustrate the huge amount of investment that has gone into land administration and property formalization and therefore the wealth of experience in the region. In the following section of this paper I bring this experience to bear on certain key land administration issues, namely the question of costs.

2. COMPARATIVE ANALYSIS OF COSTS

While significant resources have been invested by the donor community in modernizing land administration infrastructure around the globe, there has been little systematic discussion and documentation of actual costs. Better understanding of the underlying issues and the trade-offs involved in choosing among different technical, legal and institutional options of providing land administration services is needed. Even though the World Bank, IDB and other donors have long supported titling interventions all over the world, surprisingly little is known about the actual costs of such interventions, both in terms of project implementation and comparative transaction costs once the new systems are in place. Until recently, little effort has been made to disaggregate costs into the specific activities required to formalize a piece of land.
 

In reviewing previous studies that dealt with costs, there are several worth mentioning. In 1985 Janice Bernstein at the World Bank documented a study she carried out entitled “The costs of land information systems.” (Bernstein 1985) She compiled information on the topic through “a review of the literature and illustrative programs as well as discussion with experts in the field…” (p. 5) She concluded early on that “..there is a great need for coordinated research among international organizations and training institutions focusing on the economics of land information..” (p. 11) The report focused largely on the potential for lowering the cost of cadastral surveys through inertial surveying and GPS, which were just becoming operational at that time. It also focused on methods for estimating the cost of photogrammetric mapping. A fully operational GPS/GNSS system, the higher precision of today’s satellite imagery, and airborne GPS have essentially reduced the value of this information to one of historical interest. The study also contributes little towards the development of a comparative methodology.

Two years later in 1987 a symposium entitled “The Economics of Land Information” was held in Baltimore, MD, under the auspices of the Institute of Land Information (ILI 1987). This issue was topical in the US at that time as GIS was becoming mainstream and county offices were in the process of digitizing their land information. Approaches discussed at the ILI and other forums at that time included: (i) an avoided cost approach, where benefits are construed as the avoidance of downstream costs by making upstream (often public) investments in, for example, geodetic infrastructure¹ – creating the information now means that it does not have to be repeated at a later time; (ii) a ‘use and value’ approach whereby benefits are gauged relative to frequency of use – information that is used more often has more value even though it may have cost the same to produce. The ‘avoided cost’ approach may have some value, but it focuses more on future costs than more defensible present or past costs. However, both approaches are not that useful for developing a comparative methodology as they focus more on benefits than costs.

Other cross-country studies include work done by Dale and McLaughlin (1988) and Holstein (1993). Their breakdown of costs by activity is compared with that given by Bernstein (1985) in Table I below.

TABLE I. Percentage Distribution of Costs by Activity

None of these three studies provide a robust comparative analysis methodology, although they do suggest focusing on activities such as mapping, adjudication, surveying, etc. Even this can be problematic as surveying may sometimes be included as a sub-component of adjudication (Dale and McLaughlin 1990).

Gross unit costs are typically used to compare costs across different projects, without taking into account the significantly different contexts and approaches. As a result, cadastral and land registration interventions are often viewed as expensive activities that do not generate sufficient benefits to justify their costs. Furthermore, no systematic template exists for collecting data across different countries. The cost issue came to a head in a 2001 e-conference on “Lessons Learned in Land Administration” organized by the World Bank (Deininger 2003; Barnes 2003). One participant shared information on the Peruvian Titling Project (PETT) which had reduced the cost of formalizing a parcel to approximately $47 per parcel. In response to this, another participant countered that in Eastern Europe they were titling at the cost of $1.05 per parcel! Either these two participants were talking about two completely different processes and products or else the contextual setting of these two cases was incomparably different. Clearly there was a need to ‘unpack’ these numbers and develop a framework for comparing the same process or product.

Following this conference we developed a template of questions and tables that could be administered at the country level. We approached this by identifying costs at three different levels – starting from global project figures and then considering costs at the level of project components, and finally examining specific costs entailed in converting a parcel of land into a formal registered property. We also recognized the need to contextualize these studies so that the cost figures could be considered against the specific context within which the project was being implemented. Subsequently, the template was expanded to include an analysis of the effectiveness of the land administration system. This ‘template’ was then applied in seventeen countries in four different regions - Latin America and the Caribbean (El Salvador, Peru, Bolivia, Trinidad and Tobago), E. Europe and Central Asia (Armenia, Kyrgyzstan, Moldova, Latvia), Asia (Indonesia, Karnataka, Philippines, Thailand) and Africa (Ghana, Mozambique, Namibia, South Africa, Uganda). Subsequently, four regional reports were prepared that summarized the country-level reports (Barnes 2002; Adlington 2002; Land Equity 2003; Augustinus 2003). Finally a global report was prepared comparing all countries across the four regions (Land Equity 2007).

Instead of drawing on the general data for the global comparison, I have widened the LAC scope by considering 11 projects within the region. The data are drawn from the many project documents I have in my own library as well as others which are listed on the LandNet Americas portal.⁵

 TABLE II. Global Comparison of per parcel Costs - Total Project Costs/Total Parcels ⁶

This kind of comparison is of minimal use partly because it assumes that the total budget can be associated with the number of parcels that are either titled or regularized in some way. Costs associated with legal reform, institutional strengthening, equipment purchases, etc. are examples of costs that have no relation to parcels, but are still included in the comparative figures given in Table I. Although the parcel is the unit of choice when assessing the extent and cost of surveying and regularization, its cross-scalar nature produces unwelcome complexities. A ‘parcel’ may include any of the following tenure units:

• small urban lots (e.g. 20m x 30m)

• peri-urban lots

• small agricultural parcels (minifundias)

• medium rural parcels

• large rural parcels

• large communally-held parcels (e.g. indigenous communities)

The scale of a parcel may therefore vary from a small urban lot to a communal property that may approximate the size of a municipality. Additionally, at the project design stage the number of parcels in a jurisdiction or area is often the weakest data available. The whole motivation for adjudication and titling stems from the fact that there is no reliable formal parcel information in the registry or cadastre. Therefore parcel information may be available at the end of the project, but during design it can only be inferred through estimating average parcel sizes or consulting census data.

Costs are also estimated on an area (per hectare) basis? Using the project documents as a source again, those costs that are available are shown in Table I. Four out of the eleven projects in Table I do not list area to be titled or regularized in the project document. In the remainder, the per hectare costs range from $2/ha in Nicaragua to $96/ha in Panama with an average of $21/ha. This approach to unit costs suffers from the same problem as mentioned above – it assumes that the cost per unit area is uniform, ignoring the fact that the multiple scales of parcels contradict this especially in the fieldwork component. Those projects that contain several large parcels, such as indigenous communities, skew the cost per hectare numbers (such as in Nicaragua). It is therefore necessary to look deeper than these global figures if we are to effectively compare these projects. It may be more productive to consider the costs of an average size parcel.

At a more specific level we can examine costs by focusing on procurement type. Once again, this data is easily available in project documents, and the results for a small sample of countries is given in Table III below

TABLE III. Breakdown of Budgeted Costs by Procurement for Five Countries