Development and Conservation Goals in World Bank Projects

See allHide authors and affiliations

Science  19 Sep 2008:
Vol. 321, Issue 5896, pp. 1638-1639
DOI: 10.1126/science.1162756

Poverty alleviation and economic development are often thought to be incompatible with environmental protection and biodiversity conservation. Previous reviews of integrated conservation and development projects (ICDPs) have found frequent failures, leading many conservationists to conclude that such joint efforts are unwise (1, 2). Countering this pessimism is the argument that conservation can enhance economic development because undegraded nature supplies valuable ecosystem services (37).We found that World Bank projects with biodiversity goals were as successful in all development objectives, including poverty reduction and private sector development, as those that focused solely on development. Success was elusive for all types of projects: <20% of projects were rated as highly satisfactory, regardless of whether they were concerned with only development or a combination of development and biodiversity goals. The one predictor of highly satisfactory outcomes for projects involving biodiversity was the use of market mechanisms or sustainable finance approaches.

Since 1947, the World Bank has granted loans to fund over 11,000 projects, with the primary goals being economic development and poverty alleviation, but often with secondary goals that entail environment or biodiversity protection. Aid for biodiversity has grown following the 1992 Earth Summit in Rio de Janeiro (5).

As part of its project cycle, the World Bank produces Implementation Completion Reports (ICRs) in which completed projects are rated for overall performance and sustainability (8) (fig. S1). Because of their consistent format and large number, World Bank ICRs provide an untapped resource for investigating the compatibility of economic development and environmental or conservation goals. Although ICRs have weaknesses, the World Bank's comprehensive and standardized evaluation system is unrivaled, and the reports are not simply affirmations of success. Depending on the objective being scored, 15 to 40% of the projects examined here were rated as unsatisfactory or highly unsatisfactory.

To ask whether the inclusion of environmental goals hinders the success of World Bank development projects, we identified projects with ICRs completed during July 1998 to August 2006 when their format remained constant. We compared outcome scores from randomly selected projects that included both environmental and development goals to projects that lacked environmental goals (n = 97). The development-only projects were also randomly selected, but constrained so that pairs of environmental and nonenvironmental projects were matched by nation. Matching by nation ensured that any differences detected could not be attributed to geographic biases. In terms of overall performance, development-only projects scored no better than the development plus environment projects (see chart, left). If anything, development-only projects tended to more frequently earn an unsatisfactory ranking.

ICR outcome ratings for

(A) environment (orange) versus nonenvironment projects and (B) biodiversity (orange) versus nonbiodiversity projects (purple). Ratings for environment projects were slightly higher than those for nonenvironment projects (P = 0.063, Fisher's exact test). Ratings for biodiversity and nonbiodiversity projects did not differ significantly (P = 0.318, Fisher's exact test).

Projects categorized as having environmental goals included efforts aimed at irrigation and improved water quality, and thus might not be widely considered as conservation. To sharpen the analyses, we next focused on World Bank projects that explicitly included biodiversity goals. There are many fewer of these projects to sample from. In fact, we selected every available biodiversity project approved between 1993 and 2007 and with an ICR completed between July 1998 and August 2006 (n = 61). Then we randomly sampled development projects without biodiversity or broader environmental goals, again matched by nations to the extent possible (n = 51) and otherwise by region (n = 10). Comparing biodiversity and development projects to development projects, we found no evidence of a tradeoff associated with the addition of a biodiversity emphasis (see chart, page 1639, top).

ICRs also provide evaluations of more specific objectives such as environment, poverty reduction, gender equity, private sector development, and public sector management (table S1). Considering these objectives one at a time, development-only and combined development and biodiversity projects differed in only one category: environmental objectives. Only 2% of biodiversity projects had negligible environmental outcomes compared with 22% of nonbiodiversity projects. In no specific objective did biodiversity projects underperform compared with nonbiodiversity projects, including financial, poverty alleviation, or gender objectives. Thus, the data indicate no cost in terms of reduced performance if biodiversity is included as a project goal, but a significant reduction in environmental outcomes if biodiversity is not an explicit project focus. Moreover, total project costs (inflation-adjusted to 2007 values) for biodiversity projects were on average less than half those for nonbiodiversity projects (t test on log-transformed values, P < 0.001).

World Bank spending for biodiversity, broken down by ratings for achievement of poverty reduction objective.

Data are from 30 biodiversity projects that included an explicit poverty reduction objective and for which World Bank spending on biodiversity was available (9). Boxes delineate 25th to 75th percentiles. Top, middle, and bottom purple bars represent the maximum, 50th percentile, and minimum values, respectively. Filled circles represent means. Sample sizes are n = 6, 20, 3, and 1 for negligible, moderate, substantial, and high achievement of poverty reduction objectives. USD, U.S. dollars.

Although biodiversity projects did not perform more poorly than development-only projects, a substantial fraction (17.5%) were viewed as generally unsatisfactory and only 15% were assessed as having a substantial or better impact on poverty reduction. To probe biodiversity projects more deeply, we asked whether the amount spent by the World Bank on biodiversity activities affected achievement of poverty reduction objectives. Within biodiversity projects, the more money that was spent on biodiversity activities, the less likely was project success with respect to poverty reduction (ordinal regression: estimate = −8.58 × 10−8, SEM = 4.38 × 10−8, P = 0.05; see chart, below). More money spent on biodiversity may imply that social goals were relatively neglected.

The effect of sustainable financing and market mechanisms on overall success and sustainability of biodiversity projects.

(A) Project outcome is rated as highly satisfactory (orange), satisfactory (purple), or unsatisfactory (tan). (B) Project sustainability is rated as highly likely (orange), likely (purple), unlikely (tan), or highly unlikely (dark gray). Eight projects included sustainable financing and market mechanisms, and 49 projects did not.

Beyond funding, project design might influence outcome. Although it is difficult to quantify project design, a 2006 World Bank report (9) recorded the presence or absence of ten activities for each biodiversity project: (i) institution building, policies, and strategic planning; (ii) inventory, research, and monitoring; (iii) enhancing public awareness and education; (iv) protected areas; (v) enhancing production landscapes; (vi) sustainable financing and market mechanisms; (vii) nature tourism; (viii) tending to the needs of indigenous peoples; (ix) promoting biodiversity in agricultural settings; and (x) controlling invasive species. Activity profiles existed for 57 of the sampled biodiversity projects. We used ordinal regression to explore whether the presence of any of these activities was a predictor of overall success. Only sustainable financing and market mechanisms was associated with a highly satisfactory outcome [estimate, −1.848 ± 0.826 (SEM), P = 0.025] or a high likelihood of sustainability (estimate, −2.097 ± 0.810, P = 0.01; see chart, below).

Our examination of World Bank projects does not support assumptions about any negative impact of environmental or biodiversity objectives on the achievement of development objectives. This does not mean that conservation and development win-wins are easy to obtain. But it does mean that one should not assume a priori that melding conservation objectives onto development objectives is a bad idea. In addition, the only predictor of overall biodiversity project success was the presence of market mechanisms and sustainable finance. This finding supports arguments that the key to success in conservation is the development of market mechanisms and new sources of finance for conservation (1013).


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
View Abstract

Navigate This Article