Return to IRR Model, to MESAS Page, to the MESAS Resources Page, or to the Social Resources Page
Necessary Improvements in Current Resettlement Practices
How does the IRR model compare to today's mainstream practices in involuntary resettlement operations generated by development projects?
Evidence indicates that the IRR framework is in some important respects ahead of current mainstream practices,[19] and its wider adoption would significantly improve standards and performances. It brings a set of new elements, different from conventional approaches and methods. It builds upon the more advanced scholarly analyses of resettlement to date and proposes to development programs an improved way of diagnosing, costing, planning, financing, and implementing resettlement. This can substantially correct many of the current analytical flaws and implementation weaknesses, widely and legitimately criticized.
The model is fully compatible with the most advanced resettlement policies in existence today and offers a methodology capable to vastly increase consistency and effectiveness in the implementation of these policies. The World Bank's policy guidelines in resettlement, adopted first in 1980 (and formally strengthened in 1986, 1990, and 1994) have been gradually adopted, in essence, by other organizations, such as the aid agencies of the Organization for Economic Cooperation and Development (OECD) countries (OECD 1992), the Asian Development Bank (ADB) (ADB 1995), and others. Yet more institutional emphasis and incisive monitoring are required for their consistent application.[20] For instance, the economic analytical methodology employed in the preparation of projects under these guidelines is often operationally inconsistent with the guidelines themselves leaving ample room for cost externalizations and very incomplete risk analysis. The impoverishment risks analysis methodology is still to be generalized in such projects, including many co-financed by the World Bank, ADB, or OECD donor agencies. For the vast majority of developing countries, and some developed countries, which do not have any explicit policy for involuntary resettlement, the IRR model can serve as one of the building blocks for formulating such overdue policy guidelines.
As a planning and monitoring tool, the IRR has started to be used in the last four to five years in a number of projects in various countries. From India and Philippines, its use has been reported in resettlement preparation and planning (Thangaraj 1996, Spiegel 1997). Others have used the model in field supervision of resettlement operations and in project implementation monitoring work (Downing 1996a, 1996b; Sapkota 1999). For Lesotho, Scudder has developed operational indicators for measuring impoverishment risks and their actual impacts under the country's large Water Engineering project (Scudder 1999). A workshop of resettlement planners and practitioners from various states of India explored the model's planning potential, with analytical contributions reported in detail in Mathur and Marsden's book (1998).
For the most part, however, the risks of impoverishment are currently not addressed explicitly and systematically during the planning of very many projects that cause displacement. This occurs frequently in domestic projects that are not subject to in-depth and multisided screening; but to a considerable extent it has also been true in projects assisted by various bilateral or multilateral donor agencies or by credit-export entities.
The IRR model is to be used in conjunction with other analytical project tools, and it can help correct and improve some of them. We emphasize primarily the need to correct three entrenched flaws in the routine methodology of planning for such projects, flaws that account for the recurrent undertreatment of impoverishment risks. These include:
(a) The flaws and incompleteness of the conventional methods for project risk analysis;
(b) The over-reliance of project justification on the cost-benefit analysis (CBA) despite its glaring insufficiencies; and
(c) The absence of genuine consultation and involvement of the affected populations.
A few comments on each one are in order.
Flaws in Conventional Risk Methodology
Formal "risk analysis" is a subset of project economic and financial preparation methodology, and is carried out routinely. Yet the risk that displacement inflicts upon affected people are not part of the routine risk and sensitivity analyzes carried out by planners during project economic and financial analysis.
Conventional project risk analysis evaluates the sources and magnitude of risks that may adversely influence the rate of return to project investments. It estimates the switching values of key variables (such as duration of project implementation, cost overruns, availability of local co-financing) and the sensitivity of the project's net present value to possible changes in these variables. A sensitivity analysis is usually carried out for each one of these variables because they can threaten project outputs and the returns to investors. When necessary, the sensitivity tests are developed into alternative project scenarios and contingency actions, all in the effort to minimize financial risk and maximize returns to investments.
But risks to the people affected by the project are not part of the conventional economic risk analysis. While conventional project economic analysis and sensitivity tests are generally designed to identify, estimate, and help prevent the major risks to projects' rates of return and to project investors, they are not designed to also identify and measure the risks posed by the project to the other project stakeholders such as those displaced. This, quite simply, is a basic flaw in the pattern of current project risk analysis. The methodological bias is obvious: While risks to project investors, and to the invested capital, are analyzed and weighed carefully, the risks posed by the project itself to some of the project population, such as the displaced groups, are not subjected to similar rigorous, explicit, and systematic analysis. The risks resulting from displacement are only indirectly risks to investments and they remain out of the "classic" type of investment-risk analysis.[21] This conflicts with the goal of safeguarding people's interests and welfare. It conflicts also with the general policy objective of reducing poverty.
The conclusion, therefore, is that the conventional project risk analysis must be substantially reformulated and broadened, to cover risks to affected people, who are stakeholders in these projects. In other words, project risk analysis must explicitly include the risks of impoverishment highlighted in the IRR and design insurance measures, as well as, to the extent possible, risk safeguards and social safety nets.
The optimal response to predictable impoverishment risks is to search for project alternatives that could eliminate altogether the need to displace people, or could at least reduce the number of displacees. Such alternatives are sometimes technically feasible: for instance, by modifying the routing of a planned highway to circumvent existing settlements; by changing the location of a dam; or by reducing the dam's height. When it is not possible to fully avoid displacement, however, policymakers and planners are guided by the model to conceive special measures targeted against each of the predicted impoverishment risks. These measures could be of an economic, financial, technical, legal, or cultural nature. They should be commensurate with risk intensity. An experienced planner would identify which risks loom larger in each case, how they interact, and which to counteract first to achieve positive chain effects.[22]
Why Cost-Benefit Analysis is Incomplete
Further, the overreliance on cost-benefit analysis to justify projects that cause displacement is another fundamental source of mistreatment of the impoverishment risks inherent in such projects. CBA is utterly insufficient because it is only a macroeconomic tool that does not explore the distribution of either costs or benefits among project stakeholders.
The cost-benefit methodology justifies project investments by determining that the aggregate of a project's benefits outweighs the sum of project costs by an acceptable margin. But this justification is not sufficient for several reasons. First, losses and harm caused to the displaced individuals are not compensated by the aggregate benefits of development, because the distribution of these benefits is not tailored accordingly. The displacees are seldom among the direct beneficiaries of such projects.[23] Because the CBA method cannot predict and channel the allocation of a program's future benefits with reasonable certainty, the wholesale accounting of costs and benefits covers up a morally fallacious and haphazard distribution of these benefits.[24] Second, the real losses and full costs of displacement are typically not valued and measured properly. Therefore, they are not included and accounted for fully in projects' CBAs.
As the economic justification of projects is based on cost-benefit analysis, compensation levels tend to be brought down as much as possible to obey the cost-minimizing commandments of CBA. The frequent response to displacement, therefore, is to pay the least compensation possible, to externalize a large part of real costs, and to abandon the displaced people to fend for themselves after being uprooted. Even though this response has allowed impoverishment to run rampant in so many cases, it continues to be practiced widely.
The analysis of capital losses (physical, natural, human, and social capital) in the first part of this chapter documented many of the socioeconomic costs that are routinely overlooked under current procedures. A large part of these real costs is treated as "externalities" in current costing practice. Externalized out of projects' budgets, these costs are left to be borne by those who suffer the displacement. This is why the "justification" of costs to individuals through aggregate cost-benefit accounting is logically crude, and glosses over the real impoverishment impacts. It devalues individuals' losses and leaves many negative socioeconomic effects unaddressed.
The fact that development projects often produce real long-term gains for beneficiaries does not make the uprooting less painful for the displacees. In real-life terms, personal costs are neither fully subtracted from the aggregate benefits, nor paid for by the project's beneficiaries. These costs are covered only in small part by the meager compensation for expropriated assets and are borne in disproportionately large part by the population group victimized in the name of the "greater good for the greater numbers." This kind of spurious rationality conflicts with social justice. It vitiates both development philosophy and planning practice. Moreover, tolerance vis-à-vis this incomplete methodology in project economic analyses also does not encourage planners to seek genuine alternative solutions.[25] In sum, the CBA methodology, and its distorted application, are responsible for unnecessary tolerance of risks and, by omission, the magnification of perverse effects which otherwise could be counteracted.
To overcome the incompleteness of CBA, explicit distributional analysis should be introduced as mandatory in the methodology of development projects. Projects that involve displacement are a special sub-category of projects, more complex in their effects than others, and they should therefore include a special economic and financial analysis, distinct from CBA. This analysis must be focused on distribution. The "equity compass" requires that cost and benefits be calculated distinctly for each population category affected, positively or negatively (rather than indiscriminately for the society at large or for the project as a whole). This requirement flows from the principle that differential impacts must be recognized. This is, in fact, the principle that led to adopting safeguarding policies, such as the resettlement policy. This principle should therefore be translated in analytical economic methodologies able to make such differentiation.
The survival of improper methodologies for costing resettlement is due in many countries to the absence of national (domestic) policy and legal frameworks that define the rights and entitlements of people affected by state-imposed displacements or private-sector investments that cause displacement. Within such policy vacuums, arbitrariness easily sets in. Instituting equitable policies, as well as revising outdated policy provisions affecting resettlement (e.g., land acquisition, eminent domain law, and others) needs to occur through a process of open public debate on development's goals and means, so that not only procedures but also entrenched mindsets among decisionmakers, politicians, and technicians, can be influenced and modified.[26] It will take policy debates and policy decisions within many countries and development agencies to mandate changes in the way project economic analysis, risk analysis, and CBA are applied to projects entailing involuntary resettlement.
Perhaps the most damaging consequence of applying only and often distortingly, the cost-benefit analysis is the underfinancing of resettlement components in projects. If the losses to displaced people are not calculated fairly for compensation, and if the economic and financial analysis does not budget the investments needed (above simple compensation) to restart productive activities and trigger development, the resulting financing for resettlement is necessarily insufficient. Consequently, the impoverishment trends cannot be stopped.
In contrast, we argue that the response to displacement/impoverishment risks resulting from the model must be predicated not just on an "economics of compensation" but on an "economics of recovery" and development (Cernea 1999). The difference is fundamental. It requires the full internalization of resettlement costs and the allocation of growth-supporting investments, in addition to compensation. It implies an economic analysis of resettlement that goes beyond CBA and would lead to different patterns of financing resettlement.
It must be stated clearly: The cost of productively reestablishing a displaced family and a community is bound to exceed the strict market value of the physical losses imposed on that family or community. By definition, compensation alone is never sufficient for reestablishing a similar socioeconomic basis for resettlers. As long as resettlement planning will be centered on asset-compensation alone, it will not be able to achieve the policy goal of restoring and improving resettlers' livelihood.[27] This is why it is necessary to build a new economics of resettlement, transforming what now is essentially a compensation-based economics into an economics of recovery. The new economics of recovery would justify growth-enhancing investments in resettlement operations, in addition to providing compensation, to support resettlers' development. Formulating such an economics of recovery, with its set of analytical tools and measurements, is a professional challenge to economists. But ultimately, affirming and implementing a new economics is a political matter (see also Cernea 1996, 1999), part of a public policy response to hard dilemmas of development.
The conclusion is inescapable: Because government agencies use the weight of the state and the force of the law to impose expropriation and displacement, it is incumbent upon governments to enable those displaced to get back on their feet and share in the benefits made possible by their displacement.
Budgetary resources for enabling livelihood reconstruction can indeed be significantly supplemented through policy decisions for mandating that resettlers' would share in the stream of benefits from the projects they make possible. Van Wicklin (1999) has identified a vast range of options for providing such access to benefits. Such sharing is not only an equitable way of financing the true costs of reconstruction but also a necessity, given the limitations of other available resources.
Resettlers' Participation in Risk Analysis
Finally, the lack of consultation with the populations likely to be displaced during project preparation and before final decisionmaking compounds the fallacies introduced by inadequate economic analytical methods. It is correctly argued that participation through consultation with potentially affected people is indispensable for "resettlement in development mode" (Bartolome, de Wet, Mander 1999). The weak institutional capacity of state agencies for resettlement planning and implementation in many developing countries (Gill 1999) make participation of affected people even more necessary.
Information and communication between planners and resettlers is instrumental, in this respect, for early warnings and for making possible joint preventive activities. However, transparent information is still a rare occurrence. Dysfunctional communication between decisionmakers and groups affected by displacement are one of the roots of resettlement failure. As Mairal and Bergua (1996) have convincingly demonstrated, the risk-perception of would-be resettlers differs considerably from what technical experts and agencies tend to think about risks resulting from displacement. Their research has confirmed the hypothesis that agencies' failure to grasp what is socially perceived as risks has "played an essential role in the escalation of conflict in the Zaragoza dam area" in Spain.
For resettlers themselves, the predictive (warning) utility of the IRR model is that it enables them, and their organizations, be informed for conscious participation, negotiation, and adoption of coping (resource?mobilization) strategies, with lead time. Resettlers must receive information in a timely and transparent manner, understand well the impending displacement, and overcome disbelief or the tendency to denial. By forecasting the chain effects of displacement, the IRR model helps informed participation and prompts resettlers to search for alternatives, to resist inadequately prepared displacements before they occur, and to pursue their entitlements when displacement is unavoidable.
Conversely, breakdowns in information and communication tend to result in "reverse participation," i.e., in active opposition movements against development programs (Oliver-Smith 1994, Dwivedi 1997). The ill-advised position taken by some agencies, which maintain an information embargo about likely displacements and resettlers' entitlements, virtually guarantees such opposition. Withholding information, instead of participation and transparency, is often "justified" by officials to prevent panic and stress. In fact, however, this is deceptive and self-defeating. It preempts the early mobilization of resettlers in the reconstruction of their own livelihoods. Their energy is an exceptionally important factor, which even the resettlement literature has seldom highlighted.
Resettlement Research and the IRR Model
To conclude the presentation of the IRR framework, it is important to briefly outline its use in recent research practice and research literature, since this model was first formulated.
During the last several years, the IRR model has been increasingly discussed by researchers and practitioners and is currently "at work" in numerous development and research projects. A large study carried out by the Institute for Socioeconomic Development (ISED) in Orissa, India, took the IRR model as its conceptual and methodological basis in exploring resettlement processes caused by seven major projects (in dam construction, thermal plants, mining, and industry). The sample included 31 villages and 441 households with 2,274 people, selected from among 95 affected villages with 1,977 households. That study produced one of the most comprehensive and integrated surveys of displacement impacts published to date in India (Pandey and Associates 1998), practically confirming the framework under the demands of a large-scale field investigation. Its key findings are structured along the model's impoverishment risks. Another study focused on "countering the impoverishment risks," reported from India's Rengali dam (Ota 1996; see also Ota and Mohanty 1998); the study measured actual impacts of each risk variable, analyzing counter-risk measures and formulating recommendations about what needs to be done on the ground. Research on impoverishment risk and impacts was started in Lesotho at the request of the international panel monitoring the Water Engineering Project (Hitchcock, Scudder and assoc. 1999). In Nepal (Kali Gandaki Project) the application of the model in several ongoing impact evaluation resettlement studies has revealed positive experiences and produced operational recommendations (Sapkota 1999).
The theoretical implications and potential of the IRR have also been discussed during the last four to five years in several international scientific conferences that took place in Colombia, South Africa, India, the United States, and elsewhere. Several books and numerous articles have been devoted to discussion of the validity of the model, testing its applicability, or proposing various developments of its elements. A book published in 1999 by Indian resettlement scholar L. K. Mahapatra reports on an original test to which he submitted the model: The author undertook a vast secondary analysis of virtually the entire empirical research literature on resettlement published in India during the last 20 to 30 years, to explore whether or not the IRR model is validated by the findings reported in the research literature. His analysis confirmed the model. Parasuraman (1999) discussed the impoverishment risks identified by the IRR model in his book on displacements in India and concluded that "loss of land is the single most important cause of post-displacement impoverishment in India" (p. 45). M. Basu (1994) explored the linkages between the IRR model and the "basic needs" framework, emphasizing that people's basic needs are retrieved and addressed in the IRR model. Juliette Hayes derived and developed from the risks part of the IRR framework a "capital loss model" in displacement processes, with intriguing implications for further economic research on capital losses.
Several scholars proposed expanding the IRR by including other risks and losses, such as the loss of access to public services (Mathur 1998, 1999), loss of civil rights (Downing 1996a), or temporary loss of access to schooling for school-age children caught in the throes of displacement (Mahapatra 1999a, 1999b). Whether or not these or other risks should be added to the general risk model is a question deserving reflection, but the very proposals exploit the flexibility and adaptability of the model itself, which is prone to refinements. As this chapter goes to print, Robert Hackenberg's editorial article in the journal Human Organization is opening up a "public discussion" in that journal among anthropologists around the impoverishment risks and the economics of involuntary displacement and reconstruction (Hackenberg 1999).
Significantly, also, the first full-scale research project on conflict-caused refugee populations that explicitly tests and applies the IRR model has been completed by Robert Muggah (1999) in Colombia. This study generated important new findings about internally displaced refugees. It recommends policy and strategy for reestablishing or creating institutional capacities necessary for resettling refugees (Muggah 1999). Crisp, in turn, undertook a critical review of the United Nations High Commissioner for Refugees (UNHCR) program for Mozambique's huge post-conflict refugee population, in light of the IRR framework. In that "shattered society coming together again," Crisp revealed causes of successes and failures in the itinerary "from social disarticulation to social reconstruction" (Crisp 1996).
In short, the IRR is being increasingly used operationally in project preparation, appraisals, monitoring and evaluation work, in designing indicators or formulating recommendations, as well as in theory-led basic research. Further use of the risks and reconstruction model will certainly test its potential in more ways and will explore its relevance for various types of displacements and reconstruction approaches.
Having myself done considerable research and operational work on resettlement, I cannot emphasize enough the difficulties involved in actually preventing and mitigating its impoverishment risks and moral hazards. This is why forecasting impoverishment trends is crucial for adopting and implementing policies that avoid displacement and counteract undesirable outcomes when resettlement is unavoidable. Failure to acknowledge and make known the economic and cultural risks inherent in displacement only allows them to unfold unimpeded in every case. Conversely, equitable policy, plus planning, financing, and implementing resettlement with the participation of those affected, can create the premises for the improvement of resettlers' livelihoods.
Forward to References
Backward to The Basic Processes of Livelihood Reconstruction
-----------------------
[19] Within "mainstream practices" we can, of course, distinguish various levels and standards. The bulk of resettlement operations occur under various types of national (domestic) programs. Unfortunately, field research by many evaluators has concluded that under these operations standards tend to be very low and detrimental to most displacees. Resettlement operations under development projects assisted by international aid aim to achieve higher standards, formulated in the policies of multilateral development Banks, UN agencies, or bilateral donor agencies like those of OECD countries. But there is much to be changed and improved even in the methodology of this group of projects.
[20] Deficiencies in adherence to policy standards appear, especially during implementation, because implementation depends primarily on the political will and institutional capacity of borrowing governments and their project execution agencies. Such agencies often do not feel a strong "ownership" of these aid policy standards, despite project legal agreements, because of the absence of national policy standards.
[21] One kind of "social" risk that routine economic analysis does consider is defined in economics as "behavioral risk." The term refers to risks to the project, not the people. This includes risks that can be incurred by investments because of modifications in people's general behavior (e.g., behavioral changes in consumption patterns, in employment preferences, in support for the project altogether, etc.). For the safety of project investments, anticipating this kind of risks is, of course, necessary. But this is only a unidirectional consideration of the relations between people and projects. It does not add anything to the social safety of people that the project itself puts at risk.
[22] In the Philippines Batangas Port Development project, for instance, a social planner has introduced the IRR model to sharpen the perception by project staff of risk intensity and accordingly calibrate the reconstruction strategy (Spiegel 1997). He used a simple five-point Lickert scale to hypothesize the risk intensity for Batangas relocatees (i.e., low risk potential, moderately low, medium, moderately high, high) for each one of the eight risk variables: landlessness, joblessness, homelessness, etc. The goal was to tailor a comprehensive risk-response package which allocates differential resources commensurate to each risk-intensity, in ways attuned to specific circumstances in that location.
[23] Mohan Mathur gives the example of electricity generated by projects in the Singrauli region of India, which is available to people hundreds of miles away in Delhi and other north Indian cities, while those who gave up their lands for the construction of power plants often have no access to electricity. "In fact," he writes, "planners' view of projects as a means of eradicating poverty makes little sense to those who lose their lands and livelihoods … Development programs are increasingly being seen as inimical to people's interests and responsible for their worsening, not improving, situation" (Mathur 1999).
[24] The randomness and inequities of access to the "gains or pains" of development become obvious on the ground, when, by some "wheel of fortune"-as in the case of downstream irrigation development vis-à-vis upstream inundation and destruction-the program generates benefits for certain population segments while it inflicts adversity upon other population groups (unlucky enough to live upstream). Thus, the absence of distributional analysis in CBA legitimizes by default-and helps perpetuate-situations where some people share the gains, while others share the pains.
[25] I analyzed the economics of involuntary resettlement in a more detailed way elsewhere (see Cernea 1999). Also, for a detailed discussion of methodological fallacies in the economics of resettlement, see Pearce 1999.
[26] The broad changes needed "make indispensable a wide participation of the civil society in such "debate" on the viability and justifiability of projects themselves. This is precisely why the formulating and drafting of a policy and legislation needs to be converted into a process of generating wider public debate on the projects and the very patterns of development that create widespread displacement. If, after the due process outlined above, displacement becomes inevitable, resettlement must become an opportunity, a mandate for reconstructing production systems, raising standards of living, restoring community and kinship relations and minimizing the conflict with the host community" (Kothari 1995).
[27] "The key to development-oriented resettlement is to adopt a people-centered approach, not a property-compensation approach" (Serageldin 1994).
Return to IRR Model, to MESAS Page, to the MESAS Resources Page, or to the Social Resources Page
